NVIDIA H200 SXM 141 GB vs NVIDIA Tesla P100 DGXS

Comparison of NVIDIA H200 SXM 141 GB with 141 GB HBM3e and 16,896 cores vs NVIDIA Tesla P100 DGXS with 16 GB HBM2 and 3,584 cores.

Loading...

Performance Rating

H200 H200
MI325X MI325X
A100 A100

NVIDIA H200 SXM 141 GB

NVIDIA H200 SXM 141 GB

RX 7900 XTX RX 7900 XTX
Instinct MI300X Instinct MI300X
MI250 MI250

NVIDIA Tesla P100 DGXS

NVIDIA Tesla P100 DGXS

Contents:

Memory ML Performance Compute Power Architecture & Compatibility ML Software Support Clocks & Performance Power Consumption Rendering Benchmarks Additional

Memory

Memory Size

141 GB 16 GB

Memory Type

HBM3e HBM2

Memory Bandwidth

4.89 TB/s 732.2 GB/s

Memory Bus Width

6,144 бит 4,096 бит

ML Performance

FP16 (Half Precision)

267.6 TFLOPS 21.22 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

66.91 TFLOPS 10.61 TFLOPS

FP64 (Double Precision)

33.45 TFLOPS 5.304 TFLOPS

CUDA Cores

16,896 3,584

RT Cores

No No

Architecture & Compatibility

GPU Architecture

Hopper Pascal

SM (Streaming Multiprocessor)

132 56

PCIe Version

PCIe 5.0 x16 PCIe 3.0 x16

ML Software Support

CUDA Version

🔥 9.0
6.0

Clocks & Performance

Base Clock

1,500 1,328

Boost Clock

1,980 1,480

Memory Clock

1,593 715

Power Consumption

TDP/TGP

700 W
🔥 -57% 300 W

Recommended PSU

1100 W
🔥 -36% 700 W

Power Connector

8-pin EPS None

Rendering

Texture Units (TMU)

528 224

ROP

No No

L2 Cache

50 MB 4 MB

Benchmarks

MLPerf, llama2-70b-99.9 (UNSET)

3 534 tokens/s

MLPerf, llama2-70b-99.9 (fp16)

3 553 tokens/s

MLPerf, llama2-70b-99.9 (fp8)

2 444 tokens/s

MLPerf, llama3.1-405b (fp16)

40.8 tokens/s

MLPerf, llama3.1-405b (fp8)

25.3 tokens/s

MLPerf, llama3.1-8b (fp8)

5 161 tokens/s

MLPerf, deepseek-r1 (fp8)

1 113 tokens/s

MLPerf, mixtral-8x7b (fp8)

7 132 tokens/s

Additional

Slots

SXM Module No

Release Date

Nov. 18, 2024 April 5, 2016

Display Outputs

No outputs
No outputs

Renting is cheaper than buying