3 × R9700 LLM benchmarks
217 results · 777 measurements · 2 sources
Device specifications
| Variant | Memory | Memory type | Bus | Bandwidth | FP16 matrix | INT8 matrix |
|---|---|---|---|---|---|---|
| R9700 ↗ | 32 GB | GDDR6 | 256-bit | 640 GB/s | 191 TFLOPS | 383 TOPS |
Manufacturer specifications. Shared system RAM is not dedicated VRAM. Unreported values remain —; measured LLM results retain their reported configuration.
Selected reported tests
24 of 217 resultsSource-reported results, selected across reports by documentation quality. Match configurations before comparing values.
| View ReportGeneric LLM | Qwen3.8-Flash-Next uncensored fine-tune (exact checkpoint unreported) | HumanEval164 problems | 87.8 %Reported pass@1 · n = 164 | — | — | 3 × R9700Tensor parallel · 32 GB VRAM/GPUConfiguration
|
|---|---|---|---|---|---|---|
| View ReportLocalMaxxing | Qwen3.6-27B | Token speed | — | 691.17 tok/sPrefill · per request | 17.82 tok/sGeneration · per request | 3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | Qwen3.8-Flash-Next uncensored fine-tune (exact checkpoint unreported) | GSM8K200-question subset | 92.5 %Reported accuracy · n = 200 | — | — | 3 × R9700Tensor parallel · 32 GB VRAM/GPUConfiguration
|
| View ReportLocalMaxxing | Qwen3.6-27B | Token speed | — | 801.43 tok/sPrefill · per request | 16.68 tok/sGeneration · per request | 3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | Qwen3.8-Flash-Next uncensored fine-tune (exact checkpoint unreported) | HumanEval164 problems | 84.8 %Reported pass@1 · n = 164 | — | — | 3 × R9700Tensor parallel · 32 GB VRAM/GPUConfiguration
|
| View ReportLocalMaxxing | MiniMax-M2.5 | Token speed | — | 132.98 tok/sPrefill · unknown | 30.18 tok/sGeneration · peak · unknown | 3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | Qwen3.8-Flash-Next uncensored fine-tune (exact checkpoint unreported) | GSM8K200-question subset | 93.5 %Reported accuracy · n = 200 | — | — | 3 × R9700Tensor parallel · 32 GB VRAM/GPUConfiguration
|
| View ReportLocalMaxxing | MiniMax-M2.5 | Token speed | — | 132 tok/sPrefill · unknown | 30.39 tok/sGeneration · peak · unknown | 3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | Qwen3.8-Flash-Next uncensored fine-tune (exact checkpoint unreported) | Token speed | ~191 tok/sThroughput · aggregate | — | — | 3 × R9700Tensor parallel · 32 GB VRAM/GPUConfiguration
|
| View ReportLocalMaxxing | MiniMax-M2.5 | Token speed | — | 85.55 tok/sPrefill · unknown | 32.4 tok/sGeneration · unknown | 3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | Qwen3.8-Flash-Next uncensored fine-tune (exact checkpoint unreported) | Token speed | ~309 tok/sThroughput · aggregate | — | — | 3 × R9700Tensor parallel · 32 GB VRAM/GPUConfiguration
|
| View ReportLocalMaxxing | MiniMax-M2.5 | Token speed | — | 80.87 tok/sPrefill · unknown | 32.23 tok/sGeneration · unknown | 3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | Qwen3.8-Flash-Next uncensored fine-tune (exact checkpoint unreported) | Token speed | ~353 tok/sThroughput · aggregate | — | — | 3 × R9700Tensor parallel · 32 GB VRAM/GPUConfiguration
|
| View ReportLocalMaxxing | MiniMax-M2.5 | Token speed | — | 133.43 tok/sPrefill · unknown | 29.17 tok/sGeneration · peak · unknown | 3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | Qwen3.8-Flash-Next uncensored fine-tune (exact checkpoint unreported) | Token speed | — | — | ~42.7 tok/sGeneration · per request | 3 × R9700Tensor parallel · 32 GB VRAM/GPUConfiguration
|
| View ReportLocalMaxxing | DeepSeek-V4-Flash | Token speed | — | 269.68 tok/sPrefill · unknown | 30.78 tok/sGeneration · unknown | 3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | Qwen3.8-Flash-Next uncensored fine-tune (exact checkpoint unreported) | Token speed | ~182 tok/sThroughput · aggregate | — | — | 3 × R9700Tensor parallel · 32 GB VRAM/GPUConfiguration
|
| View ReportLocalMaxxing | DeepSeek-V4-Flash | Token speed | — | 274.29 tok/sPrefill · unknown | 30.93 tok/sGeneration · peak · unknown | 3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | Qwen3.8-Flash-Next uncensored fine-tune (exact checkpoint unreported) | Token speed | ~288 tok/sThroughput · aggregate | — | — | 3 × R9700Tensor parallel · 32 GB VRAM/GPUConfiguration
|
| View ReportLocalMaxxing | DeepSeek-V4-Flash | Token speed | — | 224.27 tok/sPrefill · unknown | 7.15 tok/sGeneration · unknown | 3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | Qwen3.8-Flash-Next uncensored fine-tune (exact checkpoint unreported) | Token speed | ~395 tok/sThroughput · aggregate | — | — | 3 × R9700Tensor parallel · 32 GB VRAM/GPUConfiguration
|
| View ReportLocalMaxxing | DeepSeek-V4-Flash | Token speed | — | 265.43 tok/sPrefill · unknown | 6.98 tok/sGeneration · unknown | 3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | Qwen3.8-Flash-Next uncensored fine-tune (exact checkpoint unreported) | Token speed | — | — | ~43.5 tok/sGeneration · per request | 3 × R9700Tensor parallel · 32 GB VRAM/GPUConfiguration
|
| View ReportLocalMaxxing | DeepSeek-V4-Flash | Token speed | — | 242.24 tok/sPrefill · unknown | 7.11 tok/sGeneration · unknown | 3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
Models and software in this collection
Model families: DeepSeek V4, GLM 4.5, GLM 4.7, GPT-OSS, Gemma 4, Laguna, Muse Glimmer, Qwen2.5, Qwen3, Qwen3.5, Qwen3.6, Qwen3.8
Engines: Radiance standalone, custom, hipfire, llama.cpp, lucebox, vllm
Methodology and evidence standards · Download the research dataset