GeekRadar / The comparison desk
COMPARE.
GPUNVIDIA H200 NVL
Graphics processor
- Video memory
- 141 GB*
- Memory bus
- 6,144 bit*
- Rated memory bandwidth
- 4,890 GB/s*
Rendering performance
Three-scene Blender rendering throughput. Versions, operating systems and rendering backends remain separate. This is not a gaming score.
Community submissions; cooling, power limits and system configuration can affect results.
Benchmark results
- LLM Inference: MLPerf: deepseek-r1 (fp4) · version unreported · score2,327.47* tokens/sPlatform unreported Compare results
- LLM Inference: MLPerf: llama2-70b-99 (fp4) · version unreported · score7,436.40* tokens/sPlatform unreported Compare results
- LLM Inference: MLPerf: llama2-70b-99.9 (fp4) · version unreported · score7,436.40* tokens/sPlatform unreported Compare results
- LLM Inference: MLPerf: llama3.1-405b (fp4) · version unreported · score130.53* tokens/sPlatform unreported Compare results
Recorded results
Only the same benchmark, version, workload, scenario and unit share a row. Missing results are not zero. System results include the full submitted hardware and software configuration.
Blender Cycles / classroomOPTIX / Linux · 4.5.0 · samples/min
NVIDIA H200 NVL
1,389.81 samples/min
9 submissions · median result
Latest submission:- Rendering backend
- OPTIX
- Operating system
- Linux
- Statistic
- Median
- Evaluation type
- community rendering submissions
Blender Cycles / Complete suiteOPTIX / Linux · 4.5.0 · samples/min
NVIDIA H200 NVL
5,760.19 samples/min
9 submissions · median result
Latest submission:- Rendering backend
- OPTIX
- Operating system
- Linux
- Statistic
- Median
- Evaluation type
- community rendering submissions
- 25th percentile
- 5718.0986412692
- 75th percentile
- 5761.548913242
Blender Cycles / junkshopOPTIX / Linux · 4.5.0 · samples/min
NVIDIA H200 NVL
1,904.64 samples/min
9 submissions · median result
Latest submission:- Rendering backend
- OPTIX
- Operating system
- Linux
- Statistic
- Median
- Evaluation type
- community rendering submissions
Blender Cycles / monsterOPTIX / Linux · 4.5.0 · samples/min
NVIDIA H200 NVL
2,459.54 samples/min
9 submissions · median result
Latest submission:- Rendering backend
- OPTIX
- Operating system
- Linux
- Statistic
- Median
- Evaluation type
- community rendering submissions
Blender Cycles / classroomOPTIX / Linux · 4.3.0 · samples/min
NVIDIA H200 NVL
1,379.77 samples/min
3 submissions · median result
Latest submission:- Rendering backend
- OPTIX
- Operating system
- Linux
- Statistic
- Median
- Evaluation type
- community rendering submissions
Blender Cycles / Complete suiteOPTIX / Linux · 4.3.0 · samples/min
NVIDIA H200 NVL
5,552.04 samples/min
3 submissions · median result
Latest submission:- Rendering backend
- OPTIX
- Operating system
- Linux
- Statistic
- Median
- Evaluation type
- community rendering submissions
- 25th percentile
- 5549.3494006733
- 75th percentile
- 5564.5454690876
Blender Cycles / junkshopOPTIX / Linux · 4.3.0 · samples/min
NVIDIA H200 NVL
1,871.98 samples/min
3 submissions · median result
Latest submission:- Rendering backend
- OPTIX
- Operating system
- Linux
- Statistic
- Median
- Evaluation type
- community rendering submissions
Blender Cycles / monsterOPTIX / Linux · 4.3.0 · samples/min
NVIDIA H200 NVL
2,298.08 samples/min
3 submissions · median result
Latest submission:- Rendering backend
- OPTIX
- Operating system
- Linux
- Statistic
- Median
- Evaluation type
- community rendering submissions
Processing & platform
- Architecture
- Hopper
- Manufacturing process
- 5 nm
- Process node
- 5 nm
- GPU cores
- 16,896*
- Graphics interface
- PCIe 5.0 x16*
- Reported chip die area
- 814 mm²*
- Transistors
- 80,000 million*
- GPU L2 cache
- 60 MB*
- Graphics processor
- GH100*
- Texture mapping units
- 528*
- Raster output units
- 24*
- Streaming multiprocessors
- 132*
- Tensor cores
- 528*
- Chip foundry
- TSMC*
- Card or module format
- Dual-slot*
- Declared GPU TDP
- 600 W*
- Suggested system PSU
- 1,000 W*
- External power connectors
- 8-pin EPS*
- OpenCL
- 3.0*
- CUDA compute capability
- 9.0*
Memory & storage
- Memory options
- 141*
- Video memory
- 141 GB*
- Video memory type
- HBM3e*
- Memory bus
- 6,144 bit*
- Rated memory bandwidth
- 4,890 GB/s*
Power specifications
- Thermal design power
- 600 W*
System configuration
Detailed specifications
Specifications 1 · 24 fields
- GPU architecture — NVIDIA H200 NVL — 141 GB HBM3e
- Hopper*
- Calculated theoretical peak: BF16 tensor dense per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 835.5* TFLOPS
Scope and conditions
- Calculation
- 1671 / 2*
- Theoretical peak: BF16 tensor with sparsity per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 1,671* TFLOPS
- Calculated theoretical peak: FP16 tensor dense per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 835.5* TFLOPS
Scope and conditions
- Calculation
- 1671 / 2*
- Theoretical peak: FP16 tensor with sparsity per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 1,671* TFLOPS
- Theoretical peak: FP32 non-tensor dense per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 60* TFLOPS
- Native hardware support: FP4 tensor — NVIDIA H200 NVL — 141 GB HBM3e
- No*
- Theoretical peak: FP64 non-tensor dense per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 30* TFLOPS
- Theoretical peak: FP64 tensor dense per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 60* TFLOPS
- Native hardware support: FP6 tensor — NVIDIA H200 NVL — 141 GB HBM3e
- No*
- Calculated theoretical peak: FP8 tensor dense per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 1,670.5* TFLOPS
Scope and conditions
- Calculation
- 3341 / 2*
- Theoretical peak: FP8 tensor with sparsity per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 3,341* TFLOPS
- Calculated theoretical peak: INT8 tensor dense per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 1,670.5* TOPS
Scope and conditions
- Calculation
- 3341 / 2*
- Theoretical peak: INT8 tensor with sparsity per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 3,341* TOPS
- Calculated theoretical peak: TF32 tensor dense per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 417.5* TFLOPS
Scope and conditions
- Calculation
- 835 / 2*
- Theoretical peak: TF32 tensor with sparsity per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 835* TFLOPS
- Physical configuration — NVIDIA H200 NVL — 141 GB HBM3e
- NVL (PCIe dual-slot, NVLink bridge)*
- GPU interconnect bandwidth (NVLink bridge (2- or 4-way)) — NVIDIA H200 NVL — 141 GB HBM3e
- 900* GB/s
- PCIe interface declaration — NVIDIA H200 NVL — 141 GB HBM3e
- PCIe Gen5: 128GB/s*
- Maximum board power — NVIDIA H200 NVL — 141 GB HBM3e
- 600* W
- Peak memory bandwidth per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 4,800* GB/s
- Memory capacity per GPU — NVIDIA H200 NVL — 141 GB HBM3e
- 141* GB
- Memory technology — NVIDIA H200 NVL — 141 GB HBM3e
- HBM3e*
- GPU model — NVIDIA H200 NVL — 141 GB HBM3e
- NVIDIA H200 NVL*
Specifications 2 · 7 fields
- Form factor declaration
- PCIe (dual-slot)*
- Graphics model
- H200 NVL*
- Full graphics designation
- NVIDIA H200 NVL*
- Manufacturer
- NVIDIA*
- Memory capacity declaration
- 141* GB
Scope and conditions
- native_context
- Capacity unit base
- Not independently established*
- Memory technology
- HBM3E*
- Thermal design power declaration
- 600* W
Specifications 3 · 42 fields
- proc transistors
- 80,000 million*
- proc die size
- 814* mm²
- cores
- 16,896*
- tensor cores
- 528*
- sm count
- 132*
- base clock
- 1,365*
- boost clock
- 1,785*
- memory clock
- 1,593*
- ram
- 141*
- name
- NVIDIA H200 NVL*
- test name Graphics benchmark result
- LLM Inference: MLPerf: llama3.1-405b (fp4)*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- LLM Inference: MLPerf: llama2-70b-99.9 (fp4)*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- LLM Inference: MLPerf: deepseek-r1 (fp4)*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- test name Graphics benchmark result
- LLM Inference: MLPerf: llama2-70b-99 (fp4)*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- ram type
- HBM3e*
- bus width
- 6,144*
- bus interface
- PCIe 5.0 x16*
- ram bandwidth
- 4.89 TB/s*
- fp16 performance
- 241.3*
- fp32 performance
- 60.32*
- fp64 performance
- 30.16*
- score Graphics benchmark result
- 2,327.472222222222*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 7,436.4*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 130.53*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- score Graphics benchmark result
- 7,436.4*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- tdp
- 600* W
- suggested psu
- 1,000* W
- slot width
- Dual-slot*
- power connectors
- 8-pin EPS*
- api cuda
- 9*
- is retail board
- No*
- gpi value
- 63.64*
- unit Graphics benchmark result
- tokens/s*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- tokens/s*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- tokens/s*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- unit Graphics benchmark result
- tokens/s*
Scope and conditions
- benchmark_context
- suite
- Per-row third-party test name*
- test_configuration_verified
- No*
- comparable
- No*
- vendor
- nvd*
- manufacturer
- NVIDIA*
- arch name
- Hopper*
- card release date
- 2024-11-18*
- proc foundry
- TSMC*
- proc process size
- 5* nm
Specifications 4 · 35 fields
- CUDA compute capability
- 9.0*
- OpenCL
- 3.0*
- Reported theoretical BF16 peak
- 965.1* TFLOPS
Scope and conditions
- Arithmetic density
- Not stated*
- Graphics interface
- PCIe 5.0 x16*
- Memory bus
- 6,144* bit
- Reported shader processors
- 16,896*
- Reported theoretical FP16 peak
- 120.6* TFLOPS
Scope and conditions
- Arithmetic density
- Not stated*
- Reported theoretical FP32 peak
- 60.32* TFLOPS
Scope and conditions
- Arithmetic density
- Not stated*
- Reported theoretical FP64 peak
- 30.16* TFLOPS
Scope and conditions
- Arithmetic density
- Not stated*
- Declared L1 cache configuration
- 256 KB (per SM)*
- Declared GPU L2 cache
- 60 MB*
- Reported display outputs
- No outputs*
- Theoretical pixel fill rate
- 42.84 GPixel/s*
- External power connectors
- 8-pin EPS*
- Reported chip die area
- 814 mm²*
- Chip foundry
- TSMC*
- Process node declaration
- 5 nm*
- Transistors
- 80.0 billion*
- Graphics processor
- GH100*
- Reported video memory
- 141* GB
- Rated memory bandwidth
- 4.89 TB/s*
- Video memory type
- HBM3e*
- Raster output units
- 24*
- Reported shading units
- 16,896*
- Card or module format
- Dual-slot*
- Streaming multiprocessors
- 132*
- Suggested system PSU
- 1000 W*
- Declared GPU TDP
- 600 W*
- Tensor cores
- 528*
- Theoretical texture fill rate
- 942.5 GTexel/s*
- Reported theoretical TF32 peak
- 482.5* TFLOPS
Scope and conditions
- Arithmetic density
- Not stated*
- Texture mapping units
- 528*
- GPU cores
- 16,896*
- Video memory
- 141* GB
- Video memory type
- HBM3e*
Specifications 5 · 1 fields
- Nominal GPU Memory
- 141* GB · HBM3e*
Scope and conditions
- native_context
- Memory scope
- Nominal per-GPU memory; not the total CPU + GPU module memory or runtime-available capacity*
- memory role
- nominal_release_inventory*
- not dedicated VRAM for shared memory systems
- Yes*
- not runtime available or HGX usable capacity
- Yes*
0 selected Compare 0
* Provisional data