AMD Radeon Instinct MI250 vs AMD Radeon R7 M260

Comparison AMD Radeon Instinct MI250 with 128 GB HBM2e and 13,312 cores vs AMD Radeon R7 M260 with 2 GB DDR3 and 384 cores.

Loading...

Performance Rating

AMD Radeon Instinct MI250 outperforms AMD Radeon R7 M260 by 8,381.67% 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

AMD Radeon R7 M260

0.6

AMD Radeon R7 M260

0.6

Expert Comparison

AMD Radeon Instinct MI250 значительно превосходит Radeon R7 M260 по всем параметрам. MI250 имеет гораздо больше ядер (13312 против 384), больше памяти (128 ГБ HBM2e против 2 ГБ DDR3) и большую пропускную способность (3.28 ТБ/с против 14.4 ГБ/с). Также MI250 обладает значительно большей производительностью в FP32 (45.26 ТФЛОПС против 0.7526 ТФЛОПС) и более высокой TDP (500 Вт).

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

Memory Type

HBM2e DDR3

Memory Bandwidth

🔥 3.28 TB/s
14.40 GB/s

Memory Bus Width

8,192 бит 64 бит

ML Performance

FP16 (Half Precision)

🔥 +48,013% 362.1 TFLOPS
0.7526 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

🔥 +5,914% 45.26 TFLOPS
0.7526 TFLOPS

FP64 (Double Precision)

🔥 +96,198% 45.26 TFLOPS
0.047 TFLOPS

CUDA Cores

🔥 +3,367% 13,312
384

RT Cores

No No

Architecture & Compatibility

GPU Architecture

CDNA 2.0 GCN 3.0

SM (Streaming Multiprocessor)

No No

PCIe Version

PCIe 4.0 x16 PCIe 3.0 x8

ML Software Support

CUDA Version

No No

CUDA Toolkit status

Supported Supported

Clocks & Performance

Base Clock

🔥 +6% 1,000
940

Boost Clock

🔥 +73% 1,700
980

Memory Clock

🔥 +78% 1,600
900

Power Consumption

Recommended PSU

900 W No

Power Connector

2x 8-pin No

TDP/TGP

500 W No

Rendering

Texture Units (TMU)

🔥 +3,367% 832
24

ROP

No No

L2 Cache

🔥 16 MB
128 KB

Benchmarks

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

63.9 tokens/s

Additional

Slots

Dual-slot No

Release Date

Nov. 8, 2021 June 11, 2014

Display Outputs

No outputs
No

Renting is cheaper than buying