NVIDIA H200 SXM 141 GB vs PNY A100 Liquid Cooled

Comparison of NVIDIA H200 SXM 141 GB with 141 GB HBM3e and 16,896 cores vs PNY A100 Liquid Cooled with 80 GB HBM2e and 6,912 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

PNY A100 Liquid Cooled

PNY A100 Liquid Cooled

Contents:

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

Memory

Memory Size

141 GB 80 GB

Memory Type

HBM3e HBM2e

Memory Bandwidth

4.89 TB/s 1.94 TB/s

Memory Bus Width

6,144 бит 5,120 бит

ML Performance

FP16 (Half Precision)

267.6 TFLOPS 77.97 TFLOPS

BF16 (Brain Float)

No 311.84 TFLOPS

TF32 (TensorFloat)

No 155.92

Compute Power

FP32 (Single Precision)

66.91 TFLOPS 19.49 TFLOPS

FP64 (Double Precision)

33.45 TFLOPS 9.746 TFLOPS

CUDA Cores

16,896 6,912

RT Cores

No No

Architecture & Compatibility

GPU Architecture

Hopper Ampere

SM (Streaming Multiprocessor)

132 108

PCIe Version

PCIe 5.0 x16 PCIe 4.0 x16

ML Software Support

CUDA Version

🔥 9.0
8.0

Clocks & Performance

Base Clock

1,500 1,065

Boost Clock

1,980 1,410

Memory Clock

1,593 1,512

Power Consumption

TDP/TGP

700 W
🔥 -57% 300 W

Recommended PSU

1100 W
🔥 -36% 700 W

Power Connector

8-pin EPS 8-pin EPS

Rendering

Texture Units (TMU)

528 432

ROP

No No

L2 Cache

50 MB 80 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 Single-slot

Release Date

Nov. 18, 2024 June 28, 2021

Display Outputs

No outputs
No outputs

Renting is cheaper than buying