NVIDIA H200 SXM 141 GB vs PC Partner RTX A2000 Embedded

Comparison of NVIDIA H200 SXM 141 GB with 141 GB HBM3e and 16,896 cores vs PC Partner RTX A2000 Embedded with 4 GB GDDR6 and 2,560 cores.

Loading...

Performance Rating

H200 H200
MI325X MI325X
A100 A100

NVIDIA H200 SXM 141 GB

67.4

NVIDIA H200 SXM 141 GB

67.4
MI250 MI250
Instinct MI300X Instinct MI300X
RX 7900 XTX RX 7900 XTX

PC Partner RTX A2000 Embedded

PC Partner RTX A2000 Embedded

Contents:

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

Memory

Memory Size

🔥 +3,425% 141 ГБ
4 GB

Memory Type

HBM3e GDDR6

Memory Bandwidth

🔥 4.89 TB/s
192.0 GB/s

Memory Bus Width

6,144 бит 128 бит

ML Performance

FP16 (Half Precision)

🔥 +3,268% 267.6 TFLOPS
7.946 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

🔥 +742% 66.91 TFLOPS
7.946 TFLOPS

FP64 (Double Precision)

🔥 +26,832% 33.45 TFLOPS
0.1242 TFLOPS

CUDA Cores

🔥 +560% 16,896
2,560

RT Cores

No 20

Architecture & Compatibility

GPU Architecture

Hopper Ampere

SM (Streaming Multiprocessor)

🔥 +560% 132
20

PCIe Version

PCIe 5.0 x16 PCIe 4.0 x8

ML Software Support

CUDA Version

🔥 9.0
8.6

Clocks & Performance

Base Clock

🔥 +38% 1,500
1,087

Boost Clock

🔥 +28% 1,980
1,552

Memory Clock

🔥 +6% 1,593
1,500

Power Consumption

TDP/TGP

700 W
🔥 -95% 35 W

Recommended PSU

1100 W No

Power Connector

8-pin EPS None

Rendering

Texture Units (TMU)

🔥 +560% 528
80

ROP

No 20

L2 Cache

🔥 +2,400% 50 MB
2 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
MXM Module

Release Date

Nov. 18, 2024 March 30, 2022

Display Outputs

No outputs
Portable Device Dependent

Renting is cheaper than buying