NVIDIA H200 SXM 141 GB vs XFX GT 610 2 GB

Comparison of NVIDIA H200 SXM 141 GB with 141 GB HBM3e and 16,896 cores vs XFX GT 610 2 GB with 2 GB DDR3 and 48 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

XFX GT 610 2 GB

XFX GT 610 2 GB

Contents:

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

Memory

Memory Size

141 GB 2 GB

Memory Type

HBM3e DDR3

Memory Bandwidth

4.89 TB/s 8.528 GB/s

Memory Bus Width

6,144 бит 64 бит

ML Performance

FP16 (Half Precision)

267.6 TFLOPS No

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

66.91 TFLOPS 0.1555 TFLOPS

FP64 (Double Precision)

33.45 TFLOPS 0.013 TFLOPS

CUDA Cores

16,896 48

RT Cores

No No

Architecture & Compatibility

GPU Architecture

Hopper Fermi 2.0

SM (Streaming Multiprocessor)

132 1

PCIe Version

PCIe 5.0 x16 PCIe 2.0 x16

ML Software Support

CUDA Version

🔥 9.0
2.1

Clocks & Performance

Base Clock

1,500 No

Boost Clock

1,980 No

Memory Clock

1,593 533

Power Consumption

TDP/TGP

700 W
🔥 -96% 29 W

Recommended PSU

1100 W
🔥 -82% 200 W

Power Connector

8-pin EPS None

Rendering

Texture Units (TMU)

528 8

ROP

No No

L2 Cache

50 MB 128 KB

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 July 5, 2012

Display Outputs

No outputs
1x DVI
1x HDMI 1.3a
1x VGA

Renting is cheaper than buying