AMD Radeon Instinct MI250 vs NVIDIA NVS 810

Comparison AMD Radeon Instinct MI250 with 128 GB HBM2e and 13,312 cores vs NVIDIA NVS 810 with 2 GB DDR3 and 512 cores.

Loading...

Performance Rating

AMD Radeon Instinct MI250 outperforms NVIDIA NVS 810 by 6,596.05% in the overall GPU ARK performance rating

A100 A100
H200 H200
MI325X MI325X

AMD Radeon Instinct MI250

50.9

AMD Radeon Instinct MI250

50.9
RX 7900 XTX RX 7900 XTX
MI250 MI250
Instinct MI300X Instinct MI300X

NVIDIA NVS 810

0.8

NVIDIA NVS 810

0.8

Expert Comparison

AMD Radeon Instinct MI250 значительно превосходит NVIDIA NVS 810 по всем ключевым параметрам. MI250 имеет гораздо больше ядер и памяти, более высокую пропускную способность и большую TDP, что делает ее идеальной для вычислений и научных расчетов. NVS 810 уступает в вычислительной мощности и эффективности, подходит для базовых задач в офисе или обучения.

Contents:

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

Memory

Memory Size

🔥 +6,300% 128 ГБ
2 GB ×2 (4 ГБ)

Memory Type

HBM2e DDR3

Memory Bandwidth

🔥 3.28 TB/s
14.40 GB/s ×2 (28.8 GB/s)

Memory Bus Width

8,192 бит 64 бит ×2 (128 бит)

ML Performance

FP16 (Half Precision)

🔥 362.1 TFLOPS
No

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

🔥 +4,179% 45.26 TFLOPS
1.0578 TFLOPS

FP64 (Double Precision)

🔥 +136,637% 45.26 TFLOPS
0.0331 TFLOPS

CUDA Cores

🔥 +2,500% 13,312
512 ×2 (1024)

RT Cores

No No

Architecture & Compatibility

GPU Architecture

CDNA 2.0 Maxwell

SM (Streaming Multiprocessor)

No No

PCIe Version

PCIe 4.0 x16 PCIe 3.0 x16

ML Software Support

CUDA Version

No 5.0

CUDA Toolkit (first supported)

v6

CUDA Toolkit (deprecated)

v12

CUDA Toolkit (support removed)

v13

CUDA Toolkit status

Supported Removed

Clocks & Performance

Base Clock

🔥 +11% 1,000
902

Boost Clock

🔥 +65% 1,700
1,033

Memory Clock

🔥 +78% 1,600
900

Power Consumption

Recommended PSU

900 W
🔥 -72% 250 W

Power Connector

2x 8-pin None

TDP/TGP

500 W
🔥 -86% 68 W

Rendering

Texture Units (TMU)

🔥 +2,500% 832
32 ×2 (64)

ROP

No No

L2 Cache

🔥 16 MB
1024 KB

Benchmarks

llama.cpp, llama-2-7b-Q4_0

63.9 tokens/s

Additional

Slots

Dual-slot Single-slot

Release Date

Nov. 8, 2021 Nov. 4, 2015

Display Outputs

No outputs
8x mini-DisplayPort 1.2

Renting is cheaper than buying