AFOX RTX 3060 vs AMD Radeon Instinct MI325X

Comparison of AFOX RTX 3060 with 12 GB GDDR6 and 3,584 cores vs AMD Radeon Instinct MI325X with 288 GB HBM3e and 19,456 cores.

Loading...

Performance Rating

H200 H200
MI325X MI325X
A100 A100

AFOX RTX 3060

AFOX RTX 3060

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

AMD Radeon Instinct MI325X

AMD Radeon Instinct MI325X

Contents:

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

Memory

Memory Size

12 GB 288 GB

Memory Type

GDDR6 HBM3e

Memory Bandwidth

360.0 GB/s 10.3 TB/s

Memory Bus Width

192 бит 8,192 бит

ML Performance

FP16 (Half Precision)

12.74 TFLOPS 653.7 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

12.74 TFLOPS 81.72 TFLOPS

FP64 (Double Precision)

0.199 TFLOPS 81.72 TFLOPS

CUDA Cores

3,584 19,456

RT Cores

28 No

Architecture & Compatibility

GPU Architecture

Ampere CDNA 3.0

SM (Streaming Multiprocessor)

28 No

PCIe Version

PCIe 4.0 x16 PCIe 5.0 x16

ML Software Support

CUDA Version

8.6 No

Clocks & Performance

Base Clock

1,320 1,000

Boost Clock

1,777 2,100

Memory Clock

1,875 2,525

Power Consumption

TDP/TGP

🔥 -83% 170 W
1000 W

Recommended PSU

🔥 -68% 450 W
1400 W

Power Connector

1x 8-pin None

Rendering

Texture Units (TMU)

112 1,216

ROP

28 No

L2 Cache

3 MB 16 MB

Benchmarks

MLPerf, llama2-70b-99.9 (Dummy)

3 596 tokens/s

MLPerf, llama2-70b-99.9 (fp8)

1 946 tokens/s

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

22.4 tokens/s

MLPerf, mixtral-8x7b (fp8)

6 975 tokens/s

Additional

Slots

Dual-slot OAM Module

Release Date

Jan. 12, 2021 Oct. 12, 2024

Display Outputs

1x HDMI 2.1
3x DisplayPort 1.4a
No outputs

Renting is cheaper than buying