GeekRadar / The comparison desk
COMPARE.
GPURadeon R9 FURY X2
Graphics processor
Benchmark results
- Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024 · version unreported · score1,079.59* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048 · version unreported · score964.38* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512 · version unreported · score1,138.88* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128 · version unreported · score59.44* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024 · version unreported · score1,721.44* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048 · version unreported · score1,680.48* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512 · version unreported · score1,726.09* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128 · version unreported · score71.45* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128 · version unreported · score26.29* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024 · version unreported · score236.95* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048 · version unreported · score230.31* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512 · version unreported · score237.79* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128 · version unreported · score35.79* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024 · version unreported · score1,337.74* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048 · version unreported · score1,278.18* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512 · version unreported · score1,363.67* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128 · version unreported · score25.86* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024 · version unreported · score88.56* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048 · version unreported · score83.29* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512 · version unreported · score90.28* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128 · version unreported · score37.59* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024 · version unreported · score128.17* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048 · version unreported · score124.31* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512 · version unreported · score130.11* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128 · version unreported · score40.17* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024 · version unreported · score120.17* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048 · version unreported · score113.19* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512 · version unreported · score118.80* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128 · version unreported · score14.25* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024 · version unreported · score30.14* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048 · version unreported · score28.95* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512 · version unreported · score30.19* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128 · version unreported · score19.70* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024 · version unreported · score116.44* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048 · version unreported · score113.97* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512 · version unreported · score117.26* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128 · version unreported · score14.04* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024 · version unreported · score914.59* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048 · version unreported · score834.67* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512 · version unreported · score953.45* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128 · version unreported · score70.92* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024 · version unreported · score1,427.20* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048 · version unreported · score1,416.03* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512 · version unreported · score1,419.88* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128 · version unreported · score78.56* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024 · version unreported · score1,100* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048 · version unreported · score1,060.54* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512 · version unreported · score1,129.73* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128 · version unreported · score26.17* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024 · version unreported · score224.52* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048 · version unreported · score212.58* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512 · version unreported · score225.25* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128 · version unreported · score46.20* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128 · version unreported · score27.44* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024 · version unreported · score1,098.57* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048 · version unreported · score1,030.45* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512 · version unreported · score1,114.99* Tokens Per SecondPlatform unreported Compare results
- Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128 · version unreported · score25.79* Tokens Per SecondPlatform unreported Compare results
Processing & platform
- Architecture
- GCN 3.0*
- Process node
- 28 nm*
- GPUs per card
- 2*
- GPU base clock
- 1,050 MHz*
- Reported GPU frequency
- 1,050 MHz*
- Graphics interface
- PCIe 3.0 x16*
- Reported chip die area
- 596 mm²*
- Transistors
- 8,900 million*
- Graphics processor
- Capsaicin*
- Chip foundry
- TSMC*
- Card or module format
- Dual-slot*
- Suggested system PSU
- 200 W*
- External power connectors
- 2x 8-pin*
- DirectX / feature level
- 12 (12_0)*
- OpenGL
- 4.6*
- OpenCL
- 2.1*
- Vulkan
- 1.2.170*
- Shader model
- 6.5*
- Reported video memory (allocation unspecified)
- 4 GB*
- Reported shader processors (allocation unspecified)
- 4,096*
- Reported memory bus (allocation unspecified)
- 4,096 bit*
- Reported memory bandwidth (allocation unspecified)
- 512 GB/s*
Memory & storage
- Video memory type
- Reported value 1: GDDR6*Reported value 2: HBM*Reported value 3: HBM*Reported value 4: HBM*Reported value 5: HBM*
System configuration
Detailed specifications
Specifications 1 · 16 fields
- Product Name
- Radeon R9 FURY X2*
- Product Name
- Radeon R9 FURY X2*
- GPU Chip
- Capsaicin*
- GPU Chip
- Capsaicin*
- Released
- Never Released*
- Released
- Never Released*
- Bus
- PCIe 3.0 x16*
- Bus
- PCIe 3.0 x16*
- Memory
- 4 GB, HBM, 4096 bit*
- Memory
- 4 GB, HBM, 4096 bit*
- GPU clock
- 1,050* MHz
- GPU clock
- 1,050* MHz
- Memory clock
- 500* MHz
- Memory clock
- 500* MHz
- Shaders TMUs ROPs
- 4096 / 256 / 64*
- Shaders TMUs ROPs
- 4096 / 256 / 64*
Specifications 2 · 10 fields
- Foundry
- TSMC*
- Vendor
- AMD*
- FP32 GFLOPS
- 8,602* GFLOPS
- Product
- AMD Radeon R9 FURY X2*
- Type
- GPU*
- Release Date
- NaT*
- Process Size (nm)
- 28* nm
- Die Size (mm^2)
- 596* mm²
- Transistors (million)
- 8,900*
- Freq (MHz)
- 1,050* MHz
Specifications 3 · 195 fields
- proc transistors
- 8,900 million*
- proc die size
- 596* mm²
- Reported shader processors (allocation unspecified)
- 4,096*
- memory clock
- 500*
- Reported video memory (allocation unspecified)
- 4*
- name
- AMD Radeon R9 FURY X2*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- ram type
- HBM*
- bus width
- 4,096*
- bus interface
- PCIe 3.0 x16*
- Reported memory bandwidth (allocation unspecified)
- 512* GB/s
- fp32 performance
- 8.602*
- score Graphics benchmark result
- 113.19*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,427.2*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 90.28*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 914.59*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 237.79*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,079.59*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 88.56*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,363.67*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 25.79*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 14.04*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 116.44*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 59.44*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 120.17*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,337.74*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,100*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,416.03*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 117.26*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 236.95*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 46.2*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 40.17*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 124.31*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,114.99*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 113.97*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 128.17*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 37.59*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 78.56*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 27.44*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,129.73*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 14.25*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,419.88*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 70.92*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 83.29*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 230.31*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 28.95*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 224.52*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 225.25*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 19.7*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,680.48*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 30.14*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,060.54*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,726.09*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,721.44*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 35.79*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 25.86*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 118.8*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 71.45*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 26.17*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,098.57*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,030.45*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 834.67*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,138.88*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 953.45*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 130.11*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 212.58*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 964.38*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 30.19*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 1,278.18*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 26.29*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- suggested psu
- 200* W
- slot width
- Dual-slot*
- power connectors
- 2x 8-pin*
- is retail board
- No*
- gpi value
- 8.3*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- Tokens Per Second*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- vendor
- amd*
- manufacturer
- AMD*
- arch name
- GCN 3.0*
- proc foundry
- TSMC*
- proc process size
- 28* nm
Specifications 4 · 34 fields
- DirectX / feature level
- 12 (12_0)*
- OpenCL
- 2.1*
- OpenGL
- 4.6*
- Shader model
- 6.5*
- Vulkan
- 1.2.170*
- Architecture
- GCN 3.0*
- Graphics interface
- PCIe 3.0 x16*
- Memory bus (allocation unspecified)
- 4,096* bit
- Reported shader processors (allocation unspecified)
- 4,096*
- Reported theoretical FP32 peak
- 8.602* TFLOPS
Scope and conditions
- Arithmetic density
- Not stated*
- Declared L1 cache configuration
- 16 KB (per CU)*
- Declared GPU L2 cache (allocation unspecified)
- 2 MB*
- Reported display outputs
- 1x HDMI 1.4a, 3x DisplayPort 1.2*
- Theoretical pixel fill rate
- 67.20 GPixel/s*
- External power connectors
- 2x 8-pin*
- Reported chip die area
- 596 mm²*
- Chip foundry
- TSMC*
- Process node declaration
- 28 nm*
- Transistors
- 8,900 million*
- Graphics processor
- Capsaicin*
- GPUs per card
- 2*
- Reported video memory (allocation unspecified)
- 4* GB
- Rated memory bandwidth (allocation unspecified)
- 512.0 GB/s*
- Video memory type
- HBM*
- Raster output units (allocation unspecified)
- 64*
- Reported shading units (allocation unspecified)
- 4,096*
- Card or module format
- Dual-slot*
- Suggested system PSU
- 200 W*
- 2:4 tensor sparsity
- No*
- Theoretical texture fill rate
- 268.8 GTexel/s*
- Texture mapping units (allocation unspecified)
- 256*
- Reported shader processors (allocation unspecified)
- 4,096*
- Reported video memory (allocation unspecified)
- 4* GB
- Video memory type
- HBM*
Specifications 5 · 25 fields
- manufacturer
- AMD*
- manufacturer
- AMD*
- rop
- 48*
- rop
- 64*
- igp
- No*
- igp
- No*
- bus
- PCIe 3.0 x16*
- bus
- PCIe 4.0 x8*
- memType
- GDDR6*
- memType
- HBM*
- gpuChip
- AD107*
- gpuChip
- Fiji*
- productName
- Radeon R9 FURY X2*
- productName
- Radeon R9 FURY X2*
- memSize
- 4*
- memSize
- 4*
- Reported memory bus (allocation unspecified)
- 4,096*
- gpuClock
- 1,050*
- gpuClock
- 1,830*
- memClock
- 2,125*
- memClock
- 500*
- unifiedShader
- 4,096*
- unifiedShader
- 3,072*
- tmu
- 96*
- tmu
- 256*
0 selected Compare 0
* Provisional data