AMD-Pro-BenchCommunity

13,812 benchmarks in 32 GPU combinations documented

Qwen3.8-27B-NVFP4 on R9700

51 results · 3 sources · 72 measurements

Explore all results

Grouped by reported model identity. Exact weights, quantizations and software can differ. How results are reviewed

Documented results

24 selected tests
View
ReportCustom evaluation		Qwen3.8-27B-NVFP4NVFP4 converted to MXFP4 W4A8Model ref. ↗GSM8K500 questions8 concurrent requests
97.6 %Accuracy · 488 / 500 · Reported correctness
——2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux (inferred; host distribution unreported) (inferred)
Engine
vLLM Radiance
Backend
ROCm / HIP
Harness
GSM8K evaluation (client unreported)
Prompt / output
? / ? tokens
Context
Not reported
Concurrency
8
Date
2026-09-15
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 -> MXFP4Model ref. ↗Token speed0.6.0 · chat——
139.2 tok/sGeneration · median · per request
2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux 7.2.6-x64v3-xanmod1 VM
Engine
vLLM
Backend
ROCm / HIP
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
262144
Concurrency
1
Date
2026-09-16
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4MXFP4Model ref. ↗Token speedweighted——
153.8 tok/sGeneration · mean · per request
1 × R9700Single GPU · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Not reported
Engine
Paiton + vLLM
Backend
ROCm
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
65536
Concurrency
1
Date
2026-09-26
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 converted to MXFP4 W4A8Model ref. ↗Token speed——
213.4 tok/sGeneration · per request
2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux (inferred; host distribution unreported) (inferred)
Engine
vLLM Radiance
Backend
ROCm / HIP
Harness
BetterBench
Prompt / output
? / ? tokens
Context
Not reported
Concurrency
1
Date
2026-09-15
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 -> MXFP4Model ref. ↗Token speed0.6.0 · code——
218.1 tok/sGeneration · median · per request
2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux 7.2.6-x64v3-xanmod1 VM
Engine
vLLM
Backend
ROCm / HIP
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
262144
Concurrency
1
Date
2026-09-16
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4MXFP4Model ref. ↗Token speedchat——
121.2 tok/sGeneration · mean · per request
1 × R9700Single GPU · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Not reported
Engine
Paiton + vLLM
Backend
ROCm
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
65536
Concurrency
1
Date
2026-09-26
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 converted to MXFP4 W4A8Model ref. ↗Token speed
183.9 tok/sThroughput · aggregate
——2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux (inferred; host distribution unreported) (inferred)
Engine
vLLM Radiance
Backend
ROCm / HIP
Harness
BetterBench
Prompt / output
? / ? tokens
Context
Not reported
Concurrency
1
Date
2026-09-15
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 -> MXFP4Model ref. ↗Token speed0.6.0 · file_edit——
268 tok/sGeneration · median · per request
2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux 7.2.6-x64v3-xanmod1 VM
Engine
vLLM
Backend
ROCm / HIP
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
262144
Concurrency
1
Date
2026-09-16
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4MXFP4Model ref. ↗Token speedcode——
179.9 tok/sGeneration · mean · per request
1 × R9700Single GPU · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Not reported
Engine
Paiton + vLLM
Backend
ROCm
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
65536
Concurrency
1
Date
2026-09-26
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 converted to MXFP4 W4A8Model ref. ↗Token speed
319.5 tok/sThroughput · aggregate
——2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux (inferred; host distribution unreported) (inferred)
Engine
vLLM Radiance
Backend
ROCm / HIP
Harness
BetterBench
Prompt / output
? / ? tokens
Context
Not reported
Concurrency
2
Date
2026-09-15
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 -> MXFP4Model ref. ↗Token speed0.6.0 · json——
254 tok/sGeneration · median · per request
2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux 7.2.6-x64v3-xanmod1 VM
Engine
vLLM
Backend
ROCm / HIP
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
262144
Concurrency
1
Date
2026-09-16
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4MXFP4Model ref. ↗Token speedfile edit——
179.9 tok/sGeneration · mean · per request
1 × R9700Single GPU · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Not reported
Engine
Paiton + vLLM
Backend
ROCm
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
65536
Concurrency
1
Date
2026-09-26
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 converted to MXFP4 W4A8Model ref. ↗Token speed
456.7 tok/sThroughput · aggregate
——2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux (inferred; host distribution unreported) (inferred)
Engine
vLLM Radiance
Backend
ROCm / HIP
Harness
BetterBench
Prompt / output
? / ? tokens
Context
Not reported
Concurrency
4
Date
2026-09-15
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 -> MXFP4Model ref. ↗Token speed0.6.0 · math——
252.5 tok/sGeneration · median · per request
2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux 7.2.6-x64v3-xanmod1 VM
Engine
vLLM
Backend
ROCm / HIP
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
262144
Concurrency
1
Date
2026-09-16
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4MXFP4Model ref. ↗Token speedjson——
217.5 tok/sGeneration · mean · per request
1 × R9700Single GPU · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Not reported
Engine
Paiton + vLLM
Backend
ROCm
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
65536
Concurrency
1
Date
2026-09-26
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 converted to MXFP4 W4A8Model ref. ↗Token speed
597.5 tok/sThroughput · aggregate
——2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux (inferred; host distribution unreported) (inferred)
Engine
vLLM Radiance
Backend
ROCm / HIP
Harness
BetterBench
Prompt / output
? / ? tokens
Context
Not reported
Concurrency
8
Date
2026-09-15
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 -> MXFP4Model ref. ↗Token speed0.6.0 · prose——
118.1 tok/sGeneration · median · per request
2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux 7.2.6-x64v3-xanmod1 VM
Engine
vLLM
Backend
ROCm / HIP
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
262144
Concurrency
1
Date
2026-09-16
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4MXFP4Model ref. ↗Token speedmath——
183.8 tok/sGeneration · mean · per request
1 × R9700Single GPU · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Not reported
Engine
Paiton + vLLM
Backend
ROCm
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
65536
Concurrency
1
Date
2026-09-26
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 converted to MXFP4 W4A8Model ref. ↗Token speed—
4,883 tok/sPrefill · per request
—2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux (inferred; host distribution unreported) (inferred)
Engine
vLLM Radiance
Backend
ROCm / HIP
Harness
BetterBench
Prompt / output
? / ? tokens
Context
Not reported
Concurrency
1
Date
2026-09-15
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 -> MXFP4Model ref. ↗Token speed0.6.0 · reasoning——
139.8 tok/sGeneration · median · per request
2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux 7.2.6-x64v3-xanmod1 VM
Engine
vLLM
Backend
ROCm / HIP
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
262144
Concurrency
1
Date
2026-09-16
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4MXFP4Model ref. ↗Token speedprose——
78.5 tok/sGeneration · mean · per request
1 × R9700Single GPU · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Not reported
Engine
Paiton + vLLM
Backend
ROCm
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
65536
Concurrency
1
Date
2026-09-26
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 converted to MXFP4 W4A8Model ref. ↗Token speed—
5,053 tok/sPrefill · per request
—2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux (inferred; host distribution unreported) (inferred)
Engine
vLLM Radiance
Backend
ROCm / HIP
Harness
BetterBench
Prompt / output
? / ? tokens
Context
Not reported
Concurrency
1
Date
2026-09-15
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4NVFP4 -> MXFP4Model ref. ↗Token speed0.6.0 · summarization——
198.9 tok/sGeneration · median · per request
2 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux 7.2.6-x64v3-xanmod1 VM
Engine
vLLM
Backend
ROCm / HIP
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
262144
Concurrency
1
Date
2026-09-16
Full record →
View
ReportBetterBench		Qwen3.8-27B-NVFP4MXFP4Model ref. ↗Token speedreasoning——
117.8 tok/sGeneration · mean · per request
1 × R9700Single GPU · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Not reported
Engine
Paiton + vLLM
Backend
ROCm
Harness
BetterBench 0.6.0
Prompt / output
? / ? tokens
Context
65536
Concurrency
1
Date
2026-09-26
Full record →
Quantizations and software

Quantizations: MXFP4, NVFP4 -> MXFP4, NVFP4 converted to MXFP4 W4A8
Engines: Paiton + vLLM, vLLM, vLLM Radiance

Explore all 51 results →

This is a selection across sources, not a ranking or a matched experiment. Reading benchmark reports

GPU benchmark pages: R9700