adding raw benchmark results

benchmark/generate_results.json.py

@@ -0,0 +1,239 @@
+#!/usr/bin/env python3
+import re, glob, os, json, time
+from pathlib import Path
+
+RESULTS_DIR = "results"
+OUT_JSON = "results.json"
+
+# --- Regexes ---------------------------------------------------------------
+
+# Table headers come in two shapes (with or without "fa" column)
+HEADER_RE = re.compile(r"^\|\s*model\s*\|", re.IGNORECASE)
+SEP_RE    = re.compile(r"^\|\s*-+")
+
+# Build line, e.g. "build: cd6983d5 (6119)"
+BUILD_RE  = re.compile(r"build:\s*([0-9a-f]{7,})\s*\((\d+)\)", re.IGNORECASE)
+
+# Error classifiers (same spirit as your table script)
+LOAD_ERR   = re.compile(r"failed to load model|Device memory allocation.*failed|⚠️\s*Fail", re.IGNORECASE)
+HANG_ERR   = re.compile(r"GPU Hang|HW Exception", re.IGNORECASE)
+GENERIC_ERR= re.compile(r"error:|exit \d+|runtime error|⚠️\s*Runtime Error", re.IGNORECASE)
+
+# Extract numeric ± numeric from the last column
+TS_RE      = re.compile(r"([\d.]+)\s*±\s*([\d.]+)")
+
+# Quantization from model name
+QUANT_RE = re.compile(r"(Q\d+_[A-Z_]+|BF16|F16|F32|mxfp\d+)", re.IGNORECASE)
+
+# Params like "235.09 B" from the table
+PARAMS_RE = re.compile(r"([\d.]+)\s*B", re.IGNORECASE)
+# File size like "96.99 GiB" from the table
+GIB_RE = re.compile(r"([\d.]+)\s*GiB", re.IGNORECASE)
+
+# "30B", "235B" from model name
+NAME_B_RE = re.compile(r"(\d+(?:\.\d+)?)B")
+
+# Shard suffix in filenames
+SHARD_RE = re.compile(r"-000\d+-of-000\d+", re.IGNORECASE)
+
+# --- Helpers ---------------------------------------------------------------
+
+def clean_model_name(raw):
+    base = SHARD_RE.sub("", raw)
+    return base
+
+def parse_env_and_fa(basename):
+    # pattern: <model>__<env>[__fa1]
+    parts = basename.split("__")
+    if len(parts) < 2:
+        return None, False
+    env = parts[1]
+    fa = (len(parts) > 2 and parts[2].lower() == "fa1")
+    return env, fa
+
+def env_base_and_variant(env):
+    # e.g. "rocm6_4_2-rocwmma" -> ("rocm6_4_2", "rocwmma")
+    if "-" in env:
+        base, variant = env.split("-", 1)
+        return base, variant
+    return env, None
+
+def detect_error(text):
+    if LOAD_ERR.search(text):
+        return True, "load"
+    if HANG_ERR.search(text):
+        return True, "hang"
+    if GENERIC_ERR.search(text):
+        return True, "runtime"
+    return False, None
+
+def parse_table(text):
+    """
+    Returns list of rows parsed from the markdown-like table.
+    Each row is a dict of the parsed columns, normalized by header names.
+    Handles presence/absence of the 'fa' column.
+    """
+    lines = text.splitlines()
+    rows = []
+    header = None
+    col_idx = {}
+
+    for i, line in enumerate(lines):
+        if HEADER_RE.search(line):
+            # header line
+            header = [c.strip().lower() for c in line.strip().strip("|").split("|")]
+            # next line should be the separator; skip it
+            # build index map
+            for idx, name in enumerate(header):
+                col_idx[name] = idx
+            continue
+        if header and (SEP_RE.search(line) or not line.strip()):
+            # skip separators / blanks after header
+            continue
+        if header and line.startswith("|"):
+            parts = [c.strip() for c in line.strip().strip("|").split("|")]
+            # guard for short lines
+            if len(parts) < len(header):
+                continue
+            row = {}
+            for name, idx in col_idx.items():
+                row[name] = parts[idx]
+            rows.append(row)
+        # stop parsing block when a blank line after some rows appears
+        if header and line.strip() == "" and rows:
+            break
+
+    return rows
+
+def coerce_float(m, default=None):
+    try:
+        return float(m)
+    except:
+        return default
+
+def extract_quant(model_name):
+    m = QUANT_RE.search(model_name)
+    return (m.group(1).upper() if m else None)
+
+def b_from_name(model_name):
+    m = NAME_B_RE.search(model_name)
+    return coerce_float(m.group(1)) if m else None
+
+# --- Main scan -------------------------------------------------------------
+
+runs = []
+builds = set()
+envs  = set()
+
+for path in sorted(glob.glob(os.path.join(RESULTS_DIR, "*.log"))):
+    base = os.path.basename(path).rsplit(".log", 1)[0]
+    if "__" not in base:
+        continue
+
+    model_raw, _rest = base.split("__", 1)
+    env, fa_from_name = parse_env_and_fa(base)
+    envs.add(env)
+
+    model_clean = clean_model_name(model_raw)
+
+    with open(path, errors="ignore") as f:
+        text = f.read()
+
+    # build info (take the last match in file if many)
+    build_hash, build_num = None, None
+    for m in BUILD_RE.finditer(text):
+        build_hash, build_num = m.group(1), m.group(2)
+    if build_hash:
+        builds.add((build_hash, build_num))
+
+    # detect error (if there is no valid table rows)
+    table_rows = parse_table(text)
+
+    # If table rows exist, we’ll still mark errors only if no perf found
+    has_pp = any(r.get("test","").lower()=="pp512" for r in table_rows)
+    has_tg = any(r.get("test","").lower()=="tg128" for r in table_rows)
+    error, etype = (False, None)
+    if not (has_pp or has_tg):
+        error, etype = detect_error(text)
+
+    # Determine FA flag:
+    #   prefer explicit column "fa" if present, else fallback to filename "__fa1"
+    fa_in_table = None
+    for r in table_rows:
+        if "fa" in r:
+            try:
+                fa_in_table = int(r["fa"]) == 1
+            except:
+                fa_in_table = None
+            break
+    fa_enabled = fa_in_table if fa_in_table is not None else fa_from_name
+
+    # Normalize env base / variant (e.g., rocwmma)
+    env_base, env_variant = env_base_and_variant(env)
+
+    # Emit one run per row (pp512 / tg128)
+    for r in table_rows or [{}]:
+        test = r.get("test", "").lower() if table_rows else None
+        tps_mean, tps_std = None, None
+        if table_rows:
+            ts_field = r.get("t/s", "")
+            m = TS_RE.search(ts_field)
+            if m:
+                tps_mean = coerce_float(m.group(1))
+                tps_std  = coerce_float(m.group(2))
+
+        # parse numeric helpers from row (if present)
+        params_b = None
+        file_size_gib = None
+        if "params" in r:
+            pm = PARAMS_RE.search(r["params"])
+            if pm: params_b = coerce_float(pm.group(1))
+        if "size" in r:
+            sm = GIB_RE.search(r["size"])
+            if sm: file_size_gib = coerce_float(sm.group(1))
+
+        backend = r.get("backend")
+        ngl = r.get("ngl")
+        mmap = r.get("mmap")
+
+        # quant from model name
+        quant = extract_quant(model_clean)
+        name_params_b = b_from_name(model_clean)
+
+        run = {
+            "model": model_raw,
+            "model_clean": model_clean,
+            "env": env,
+            "env_base": env_base,
+            "env_variant": env_variant,         # e.g. "rocwmma"
+            "fa": bool(fa_enabled),
+            "test": test,                       # "pp512" | "tg128" | None (if error)
+            "tps_mean": tps_mean,
+            "tps_std": tps_std,
+            "error": bool(error),
+            "error_type": etype,                # "load" | "hang" | "runtime" | None
+            "backend": backend,
+            "ngl": (int(ngl) if (ngl and ngl.isdigit()) else None),
+            "mmap": (int(mmap) if (mmap and mmap.isdigit()) else None),
+            "params_b": params_b,               # from table, if available
+            "file_size_gib": file_size_gib,     # from table, if available
+            "name_params_b": name_params_b,     # parsed from model name (e.g., 30B -> 30.0)
+            "quant": quant,
+            "log": path,
+            "build": {"hash": build_hash, "number": build_num} if build_hash else None,
+        }
+        runs.append(run)
+
+# Meta
+meta = {
+    "generated_at": time.strftime("%Y-%m-%dT%H:%M:%SZ", time.gmtime()),
+    "os_kernel": "Fedora 42 — Linux fedora 6.16.0-264.vanilla.fc42.x86_64 (2025-07-28)",
+    "llamacpp_builds": [{"hash": h, "number": n} for (h, n) in sorted(builds)],
+    "environments": sorted(envs),
+    "notes": "pp512 = prompt processing; tg128 = text generation; t/s = tokens/second",
+}
+
+out = {"meta": meta, "runs": runs}
+
+Path(OUT_JSON).write_text(json.dumps(out, indent=2))
+print(f"Wrote {OUT_JSON} with {len(runs)} rows.")

benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x2edd2a90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 failed (exit 134)

benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x432ea90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log

@@ -2,5 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-Memory access fault by GPU node-1 (Agent handle: 0x7f5e570) on address 0x7f3192c0f000. Reason: Page not present or supervisor privilege.
+| model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
-✖ ! [rocm6_4_2] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 failed (exit 134)
+| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | ROCm       | 999 |    0 |           pp512 |        129.88 ± 0.57 |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | ROCm       | 999 |    0 |           tg128 |         19.43 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+Memory access fault by GPU node-1 (Agent handle: 0x834aa90) on address 0x7f10fb96f000. Reason: Page not present or supervisor privilege.
+✖ ! [rocm6_4_2] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log

@@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-HW Exception by GPU node-1 (Agent handle: 0x16bd82e0) reason :GPU Hang
+HW Exception by GPU node-1 (Agent handle: 0x100d3790) reason :GPU Hang
 ✖ ! [rocm7_beta] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 failed (exit 134)

benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+Memory access fault by GPU node-1 (Agent handle: 0x13829790) on address 0x7fa8ef9a9000. Reason: Page not present or supervisor privilege.
+✖ ! [rocm7_beta] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log

@@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | ROCm       |  99 |    0 |           pp512 |        129.20 ± 0.38 |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | ROCm       | 999 |    0 |           pp512 |        130.17 ± 0.38 |
-| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | ROCm       |  99 |    0 |           tg128 |         19.61 ± 0.00 |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | ROCm       | 999 |    0 |           tg128 |         19.83 ± 0.00 |
 
-build: 0d883154 (6101)
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log

@@ -0,0 +1,10 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | ROCm       | 999 |  1 |    0 |           pp512 |        103.63 ± 0.10 |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | ROCm       | 999 |  1 |    0 |           tg128 |         20.09 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | Vulkan     |  99 |    0 |           pp512 |        199.54 ± 0.38 |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | Vulkan     | 999 |    0 |           pp512 |        200.76 ± 0.32 |
-| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | Vulkan     |  99 |    0 |           tg128 |         22.75 ± 0.01 |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | Vulkan     | 999 |    0 |           tg128 |         22.78 ± 0.00 |
 
-build: 0d883154 (6101)
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | Vulkan     | 999 |  1 |    0 |           pp512 |        201.86 ± 0.27 |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | Vulkan     | 999 |  1 |    0 |           tg128 |         22.83 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | Vulkan     |  99 |    0 |           pp512 |        128.00 ± 0.23 |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | Vulkan     | 999 |    0 |           pp512 |        127.73 ± 0.23 |
-| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | Vulkan     |  99 |    0 |           tg128 |         22.88 ± 0.02 |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | Vulkan     | 999 |    0 |           tg128 |         22.88 ± 0.02 |
 
-build: 0d883154 (6101)
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | Vulkan     | 999 |  1 |    0 |           pp512 |        132.54 ± 0.34 |
+| glm4moe 106B.A12B Q4_K - Medium |  68.01 GiB |   110.47 B | Vulkan     | 999 |  1 |    0 |           tg128 |         23.31 ± 0.01 |
+
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log

@@ -0,0 +1,10 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+| model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | ROCm       | 999 |    0 |           pp512 |        113.62 ± 0.21 |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | ROCm       | 999 |    0 |           tg128 |         15.47 ± 0.04 |
+
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x2f508a90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 failed (exit 134)

benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2.log

@@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices:
   Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | ROCm       |  99 |    0 |           pp512 |        124.86 ± 0.54 |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | ROCm       | 999 |    0 |           pp512 |        124.82 ± 0.18 |
-| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | ROCm       |  99 |    0 |           tg128 |         15.27 ± 0.00 |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | ROCm       | 999 |    0 |           tg128 |         15.35 ± 0.00 |
 
-build: 0d883154 (6101)
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+Memory access fault by GPU node-1 (Agent handle: 0x1527fa90) on address 0x7f55d5f6f000. Reason: Page not present or supervisor privilege.
+✖ ! [rocm6_4_2] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 failed (exit 134)

benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta.log

@@ -2,5 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-HW Exception by GPU node-1 (Agent handle: 0x2a5da2e0) reason :GPU Hang
+| model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
-✖ ! [rocm7_beta] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 failed (exit 134)
+| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | ROCm       | 999 |    0 |           pp512 |        120.54 ± 0.30 |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | ROCm       | 999 |    0 |           tg128 |         15.49 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x2a849790) reason :GPU Hang
+✖ ! [rocm7_beta] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 failed (exit 134)

benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc.log

@@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-✖ ! [rocm7_rc] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 failed (exit 134)
+| model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | ROCm       | 999 |    0 |           pp512 |        124.18 ± 0.48 |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | ROCm       | 999 |    0 |           tg128 |         15.49 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__fa1.log

@@ -0,0 +1,5 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+✖ ! [rocm7_rc] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 failed (exit 134)

benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | Vulkan     |  99 |    0 |           pp512 |        221.02 ± 0.58 |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | Vulkan     | 999 |    0 |           pp512 |        223.02 ± 0.69 |
-| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | Vulkan     |  99 |    0 |           tg128 |         16.47 ± 0.01 |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | Vulkan     | 999 |    0 |           tg128 |         16.47 ± 0.01 |
 
-build: 0d883154 (6101)
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | Vulkan     | 999 |  1 |    0 |           pp512 |        224.54 ± 0.65 |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | Vulkan     | 999 |  1 |    0 |           tg128 |         16.49 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | Vulkan     |  99 |    0 |           pp512 |        126.86 ± 0.40 |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | Vulkan     | 999 |    0 |           pp512 |        127.36 ± 0.46 |
-| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | Vulkan     |  99 |    0 |           tg128 |         16.76 ± 0.00 |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | Vulkan     | 999 |    0 |           tg128 |         16.78 ± 0.01 |
 
-build: 0d883154 (6101)
+build: cd6983d5 (6119)

benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | Vulkan     | 999 |  1 |    0 |           pp512 |        131.78 ± 0.46 |
+| glm4moe 106B.A12B Q6_K         |  94.57 GiB |   110.47 B | Vulkan     | 999 |  1 |    0 |           tg128 |         16.99 ± 0.01 |
+
+build: cd6983d5 (6119)

benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x121f0a90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 failed (exit 134)

benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x17018a90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log

@@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-HW Exception by GPU node-1 (Agent handle: 0x68b7b10) reason :GPU Hang
+HW Exception by GPU node-1 (Agent handle: 0x11442a90) reason :GPU Hang
 ✖ ! [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 failed (exit 134)

benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x64dea90) reason :GPU Hang
+✖ ! [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log

@@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-HW Exception by GPU node-1 (Agent handle: 0x1587b430) reason :GPU Hang
+HW Exception by GPU node-1 (Agent handle: 0xa636790) reason :GPU Hang
 ✖ ! [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 failed (exit 134)

benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x1417b7b0) reason :GPU Hang
+✖ ! [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log

@@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-✖ ! [rocm7_rc] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 failed (exit 134)
+| model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
+| qwen2 70B Q8_0                 |  78.21 GiB |    72.71 B | ROCm       | 999 |    0 |           pp512 |         33.30 ± 0.04 |
+| qwen2 70B Q8_0                 |  78.21 GiB |    72.71 B | ROCm       | 999 |    0 |           tg128 |          2.64 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log

@@ -0,0 +1,10 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| qwen2 70B Q8_0                 |  78.21 GiB |    72.71 B | ROCm       | 999 |  1 |    0 |           pp512 |         31.09 ± 0.02 |
+| qwen2 70B Q8_0                 |  78.21 GiB |    72.71 B | ROCm       | 999 |  1 |    0 |           tg128 |          2.65 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log

@@ -4,5 +4,5 @@ ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16:
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
 ggml_vulkan: Device memory allocation of size 2491416576 failed.
 ggml_vulkan: Requested buffer size exceeds device memory allocation limit: ErrorOutOfDeviceMemory
-main: error: failed to load model '/home/kyuz0/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf'
+main: error: failed to load model '/mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf'
 ✖ ! [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 failed (exit 1)

benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+ggml_vulkan: Device memory allocation of size 2491416576 failed.
+ggml_vulkan: Requested buffer size exceeds device memory allocation limit: ErrorOutOfDeviceMemory
+main: error: failed to load model '/mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf'
+✖ ! [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 1)

benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| qwen2 70B Q8_0                 |  78.21 GiB |    72.71 B | Vulkan     |  99 |    0 |           pp512 |         76.48 ± 0.23 |
+| qwen2 70B Q8_0                 |  78.21 GiB |    72.71 B | Vulkan     | 999 |    0 |           pp512 |         78.70 ± 0.20 |
-| qwen2 70B Q8_0                 |  78.21 GiB |    72.71 B | Vulkan     |  99 |    0 |           tg128 |          2.65 ± 0.00 |
+| qwen2 70B Q8_0                 |  78.21 GiB |    72.71 B | Vulkan     | 999 |    0 |           tg128 |          2.66 ± 0.00 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| qwen2 70B Q8_0                 |  78.21 GiB |    72.71 B | Vulkan     | 999 |  1 |    0 |           pp512 |         81.29 ± 0.14 |
+| qwen2 70B Q8_0                 |  78.21 GiB |    72.71 B | Vulkan     | 999 |  1 |    0 |           tg128 |          2.66 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0xcd80a90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 failed (exit 134)

benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x1496da90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log

@@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices:
   Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | ROCm       |  99 |    0 |           pp512 |         33.17 ± 0.07 |
+| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | ROCm       | 999 |    0 |           pp512 |         33.32 ± 0.04 |
-| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | ROCm       |  99 |    0 |           tg128 |          2.72 ± 0.00 |
+| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | ROCm       | 999 |    0 |           tg128 |          2.73 ± 0.00 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log

@@ -0,0 +1,10 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | ROCm       | 999 |  1 |    0 |           pp512 |         31.28 ± 0.02 |
+| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | ROCm       | 999 |  1 |    0 |           tg128 |          2.74 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log

@@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-HW Exception by GPU node-1 (Agent handle: 0xa5e9440) reason :GPU Hang
+HW Exception by GPU node-1 (Agent handle: 0xfeef7b0) reason :GPU Hang
 ✖ ! [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 failed (exit 134)

benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+Memory access fault by GPU node-1 (Agent handle: 0x6d017c0) on address 0x7f967f1a9000. Reason: Page not present or supervisor privilege.
+✖ ! [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log

@@ -0,0 +1,5 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+✖ ! [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | Vulkan     |  99 |    0 |           pp512 |         96.23 ± 0.16 |
+| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | Vulkan     | 999 |    0 |           pp512 |         98.14 ± 0.14 |
-| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | Vulkan     |  99 |    0 |           tg128 |          2.72 ± 0.00 |
+| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | Vulkan     | 999 |    0 |           tg128 |          2.73 ± 0.00 |
 
-build: 9c35706b (6060)
+build: cd6983d5 (6119)

benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | Vulkan     | 999 |  1 |    0 |           pp512 |         99.24 ± 0.16 |
+| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | Vulkan     | 999 |  1 |    0 |           tg128 |          2.72 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | Vulkan     |  99 |    0 |           pp512 |         79.71 ± 0.13 |
+| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | Vulkan     | 999 |    0 |           pp512 |         80.11 ± 0.09 |
-| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | Vulkan     |  99 |    0 |           tg128 |          2.72 ± 0.00 |
+| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | Vulkan     | 999 |    0 |           tg128 |          2.73 ± 0.00 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | Vulkan     | 999 |  1 |    0 |           pp512 |         82.90 ± 0.14 |
+| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | Vulkan     | 999 |  1 |    0 |           tg128 |          2.73 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x28bb9a90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x194fea90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2.log

@@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices:
   Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | ROCm       |  99 |    0 |           pp512 |        121.52 ± 0.98 |
+| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | ROCm       | 999 |    0 |           pp512 |        134.39 ± 0.32 |
-| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | ROCm       |  99 |    0 |           tg128 |         14.28 ± 0.00 |
+| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | ROCm       | 999 |    0 |           tg128 |         14.33 ± 0.00 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x3b11ea90) reason :GPU Hang
+✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta.log

@@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-HW Exception by GPU node-1 (Agent handle: 0x27159430) reason :GPU Hang
+HW Exception by GPU node-1 (Agent handle: 0x17ad57b0) reason :GPU Hang
 ✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+Memory access fault by GPU node-1 (Agent handle: 0x2314b7b0) on address 0x7f38249a9000. Reason: Page not present or supervisor privilege.
+✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc.log

@@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | ROCm       |  99 |    0 |           pp512 |        135.36 ± 0.39 |
+| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | ROCm       | 999 |    0 |           pp512 |        135.25 ± 0.50 |
-| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | ROCm       |  99 |    0 |           tg128 |         14.29 ± 0.00 |
+| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | ROCm       | 999 |    0 |           tg128 |         14.43 ± 0.00 |
 
-build: 4cb208c9 (6066)
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__fa1.log

@@ -0,0 +1,5 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | Vulkan     |  99 |    0 |           pp512 |        243.19 ± 1.20 |
+| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | Vulkan     | 999 |    0 |           pp512 |        243.45 ± 1.29 |
-| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | Vulkan     |  99 |    0 |           tg128 |         15.28 ± 0.03 |
+| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | Vulkan     | 999 |    0 |           tg128 |         15.29 ± 0.01 |
 
-build: 9c35706b (6060)
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | Vulkan     | 999 |  1 |    0 |           pp512 |        247.48 ± 1.28 |
+| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | Vulkan     | 999 |  1 |    0 |           tg128 |         15.03 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | Vulkan     |  99 |    0 |           pp512 |        137.97 ± 0.99 |
+| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | Vulkan     | 999 |    0 |           pp512 |        148.25 ± 0.91 |
-| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | Vulkan     |  99 |    0 |           tg128 |         15.07 ± 0.05 |
+| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | Vulkan     | 999 |    0 |           tg128 |         15.21 ± 0.06 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | Vulkan     | 999 |  1 |    0 |           pp512 |        149.82 ± 0.83 |
+| llama4 17Bx16E (Scout) Q6_K    |  82.35 GiB |   107.77 B | Vulkan     | 999 |  1 |    0 |           tg128 |         15.21 ± 0.04 |
+
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x9ae6a90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x6e9ba90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2.log

@@ -2,5 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-HW Exception by GPU node-1 (Agent handle: 0x2b17db10) reason :GPU Hang
+| model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
-✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 failed (exit 134)
+| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
+| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | ROCm       | 999 |    0 |           pp512 |        135.44 ± 0.76 |
+| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | ROCm       | 999 |    0 |           tg128 |         11.61 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x2fba3a90) reason :GPU Hang
+✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta.log

@@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-HW Exception by GPU node-1 (Agent handle: 0x1a77430) reason :GPU Hang
+HW Exception by GPU node-1 (Agent handle: 0x4081f7b0) reason :GPU Hang
 ✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x3c0f27b0) reason :GPU Hang
+✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__fa1.log

@@ -0,0 +1,5 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | Vulkan     |  99 |    0 |           pp512 |        238.93 ± 2.89 |
+| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | Vulkan     | 999 |    0 |           pp512 |        258.18 ± 1.38 |
-| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | Vulkan     |  99 |    0 |           tg128 |         12.25 ± 0.01 |
+| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | Vulkan     | 999 |    0 |           tg128 |         12.23 ± 0.01 |
 
-build: 9c35706b (6060)
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | Vulkan     | 999 |  1 |    0 |           pp512 |        260.16 ± 1.44 |
+| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | Vulkan     | 999 |  1 |    0 |           tg128 |         12.09 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | Vulkan     |  99 |    0 |           pp512 |        145.86 ± 2.44 |
+| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | Vulkan     | 999 |    0 |           pp512 |        168.63 ± 0.81 |
-| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | Vulkan     |  99 |    0 |           tg128 |         12.27 ± 0.00 |
+| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | Vulkan     | 999 |    0 |           tg128 |         12.26 ± 0.01 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | Vulkan     | 999 |  1 |    0 |           pp512 |        172.37 ± 0.92 |
+| llama4 17Bx16E (Scout) Q8_0    | 106.65 GiB |   107.77 B | Vulkan     | 999 |  1 |    0 |           tg128 |         12.25 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x1a40fa90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x2e0ffa90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log

@@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices:
   Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | ROCm       |  99 |    0 |           pp512 |        132.66 ± 0.56 |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | ROCm       | 999 |    0 |           pp512 |        138.27 ± 0.66 |
-| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | ROCm       |  99 |    0 |           tg128 |         17.29 ± 0.00 |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | ROCm       | 999 |    0 |           tg128 |         17.40 ± 0.00 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x3a741a90) reason :GPU Hang
+✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log

@@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | ROCm       |  99 |    0 |           pp512 |        133.71 ± 0.64 |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | ROCm       | 999 |    0 |           pp512 |        138.90 ± 0.66 |
-| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | ROCm       |  99 |    0 |           tg128 |         17.35 ± 0.00 |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | ROCm       | 999 |    0 |           tg128 |         17.62 ± 0.00 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log

@@ -0,0 +1,10 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | ROCm       | 999 |  1 |    0 |           pp512 |        123.61 ± 0.50 |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | ROCm       | 999 |  1 |    0 |           tg128 |         17.60 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log

@@ -0,0 +1,10 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | ROCm       | 999 |  1 |    0 |           pp512 |        123.58 ± 0.18 |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | ROCm       | 999 |  1 |    0 |           tg128 |         17.55 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | Vulkan     |  99 |    0 |           pp512 |        208.84 ± 1.35 |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | Vulkan     | 999 |    0 |           pp512 |        218.18 ± 0.83 |
-| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | Vulkan     |  99 |    0 |           tg128 |         20.06 ± 0.01 |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | Vulkan     | 999 |    0 |           tg128 |         20.04 ± 0.02 |
 
-build: 9c35706b (6060)
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | Vulkan     | 999 |  1 |    0 |           pp512 |        221.15 ± 0.74 |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | Vulkan     | 999 |  1 |    0 |           tg128 |         19.58 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | Vulkan     |  99 |    0 |           pp512 |        133.49 ± 1.83 |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | Vulkan     | 999 |    0 |           pp512 |        152.21 ± 0.66 |
-| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | Vulkan     |  99 |    0 |           tg128 |         19.99 ± 0.01 |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | Vulkan     | 999 |    0 |           tg128 |         19.98 ± 0.01 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | Vulkan     | 999 |  1 |    0 |           pp512 |        155.22 ± 1.09 |
+| llama4 17Bx16E (Scout) Q4_K - Medium |  57.73 GiB |   107.77 B | Vulkan     | 999 |  1 |    0 |           tg128 |         19.93 ± 0.01 |
+
+build: cd6983d5 (6119)

benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x153dfa90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 failed (exit 134)

benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+HW Exception by GPU node-1 (Agent handle: 0x2bd2ba90) reason :GPU Hang
+✖ ! [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134)

benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2.log

@@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices:
   Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | ROCm       |  99 |    0 |           pp512 |         69.48 ± 0.09 |
+| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | ROCm       | 999 |    0 |           pp512 |         74.15 ± 0.18 |
-| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | ROCm       |  99 |    0 |           tg128 |         13.54 ± 0.01 |
+| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | ROCm       | 999 |    0 |           tg128 |         13.73 ± 0.00 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+Memory access fault by GPU node-1 (Agent handle: 0x25011a90) on address 0x7fdcc1b6f000. Reason: Page not present or supervisor privilege.
+✖ ! [rocm6_4_2] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134)

benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta.log

@@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-HW Exception by GPU node-1 (Agent handle: 0x1a8d440) reason :GPU Hang
+HW Exception by GPU node-1 (Agent handle: 0x513c7b0) reason :GPU Hang
 ✖ ! [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 failed (exit 134)

benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__fa1.log

@@ -0,0 +1,6 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+Memory access fault by GPU node-1 (Agent handle: 0x2567c7c0) on address 0x7ee66236f000. Reason: Page not present or supervisor privilege.
+✖ ! [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134)

benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc.log

@@ -2,9 +2,4 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
 ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
 ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
-| model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
+✖ ! [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 failed (exit 134)
-| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | ROCm       |  99 |    0 |           pp512 |         74.69 ± 0.17 |
-| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | ROCm       |  99 |    0 |           tg128 |         13.56 ± 0.00 |
-
-build: 4cb208c9 (6066)

benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__fa1.log

@@ -0,0 +1,5 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+✖ ! [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134)

benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | Vulkan     |  99 |    0 |           pp512 |         99.94 ± 0.91 |
+| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | Vulkan     | 999 |    0 |           pp512 |        114.49 ± 0.60 |
-| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | Vulkan     |  99 |    0 |           tg128 |         15.72 ± 0.01 |
+| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | Vulkan     | 999 |    0 |           tg128 |         15.98 ± 0.01 |
 
-build: 9c35706b (6060)
+build: cd6983d5 (6119)

benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | Vulkan     | 999 |  1 |    0 |           pp512 |        116.07 ± 0.64 |
+| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | Vulkan     | 999 |  1 |    0 |           tg128 |         15.84 ± 0.01 |
+
+build: cd6983d5 (6119)

benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv.log

@@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices:
 ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | Vulkan     |  99 |    0 |           pp512 |         58.40 ± 0.21 |
+| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | Vulkan     | 999 |    0 |           pp512 |         64.85 ± 0.38 |
-| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | Vulkan     |  99 |    0 |           tg128 |         16.29 ± 0.01 |
+| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | Vulkan     | 999 |    0 |           tg128 |         16.58 ± 0.00 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv__fa1.log

@@ -0,0 +1,8 @@
+ggml_vulkan: Found 1 Vulkan devices:
+ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | Vulkan     | 999 |  1 |    0 |           pp512 |         66.76 ± 0.43 |
+| qwen3moe 235B.A22B Q3_K - Medium |  96.99 GiB |   235.09 B | Vulkan     | 999 |  1 |    0 |           tg128 |         16.83 ± 0.01 |
+
+build: cd6983d5 (6119)

benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma.log

@@ -0,0 +1,10 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+| model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
+| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       | 999 |    0 |           pp512 |        157.95 ± 2.63 |
+| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       | 999 |    0 |           tg128 |         24.53 ± 0.01 |
+
+build: cd6983d5 (6119)

benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log

@@ -0,0 +1,10 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       | 999 |  1 |    0 |           pp512 |        162.19 ± 3.06 |
+| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       | 999 |  1 |    0 |           tg128 |         24.03 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2.log

@@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices:
   Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       |  99 |    0 |           pp512 |        157.74 ± 2.65 |
+| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       | 999 |    0 |           pp512 |        157.69 ± 2.52 |
-| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       |  99 |    0 |           tg128 |         22.88 ± 0.01 |
+| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       | 999 |    0 |           tg128 |         23.89 ± 0.01 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2__fa1.log

@@ -0,0 +1,10 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       | 999 |  1 |    0 |           pp512 |        140.32 ± 2.10 |
+| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       | 999 |  1 |    0 |           tg128 |         24.33 ± 0.00 |
+
+build: cd6983d5 (6119)

benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta.log

@@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices:
   Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
 | model                          |       size |     params | backend    | ngl | mmap |            test |                  t/s |
 | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
-| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       |  99 |    0 |           pp512 |        151.25 ± 3.33 |
+| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       | 999 |    0 |           pp512 |        153.49 ± 1.19 |
-| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       |  99 |    0 |           tg128 |         23.80 ± 0.09 |
+| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       | 999 |    0 |           tg128 |         24.52 ± 0.00 |
 
-build: 66625a59 (6040)
+build: cd6983d5 (6119)

benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__fa1.log

@@ -0,0 +1,10 @@
+ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
+ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
+ggml_cuda_init: found 1 ROCm devices:
+  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
+| model                          |       size |     params | backend    | ngl | fa | mmap |            test |                  t/s |
+| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
+| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       | 999 |  1 |    0 |           pp512 |        138.49 ± 2.52 |
+| qwen3moe 30B.A3B BF16          |  56.89 GiB |    30.53 B | ROCm       | 999 |  1 |    0 |           tg128 |         24.35 ± 0.01 |
+
+build: cd6983d5 (6119)