AMD Radeon Instinct MI300X vs NVIDIA TITAN RTX

Comparison of AMD Radeon Instinct MI300X with 192 GB HBM3 and 19,456 cores vs NVIDIA TITAN RTX with 24 GB GDDR6 and 4,608 cores.

Loading...

Performance Rating

H200 H200
MI325X MI325X
A100 A100

AMD Radeon Instinct MI300X

AMD Radeon Instinct MI300X

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

NVIDIA TITAN RTX

NVIDIA TITAN RTX

Contents:

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

Memory

Memory Size

192 GB 24 GB

Memory Type

HBM3 GDDR6

Memory Bandwidth

10.3 TB/s 672.0 GB/s

Memory Bus Width

8,192 бит 384 бит

ML Performance

FP16 (Half Precision)

653.7 TFLOPS 32.62 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

81.72 TFLOPS 16.31 TFLOPS

FP64 (Double Precision)

81.72 TFLOPS 0.5098 TFLOPS

CUDA Cores

19,456 4,608

RT Cores

No 72

Architecture & Compatibility

GPU Architecture

CDNA 3.0 Turing

SM (Streaming Multiprocessor)

No 72

PCIe Version

PCIe 5.0 x16 PCIe 3.0 x16

ML Software Support

CUDA Version

No 7.5

Clocks & Performance

Base Clock

1,000 1,350

Boost Clock

2,100 1,770

Memory Clock

2,525 1,750

Power Consumption

TDP/TGP

750 W
🔥 -63% 280 W

Recommended PSU

1150 W
🔥 -48% 600 W

Power Connector

None 2x 8-pin

Rendering

Texture Units (TMU)

1,216 288

ROP

No 72

L2 Cache

16 MB 6 MB

Benchmarks

MLPerf, llama2-70b-99.9 (UNSET)

1 983 tokens/s

MLPerf, llama2-70b-99.9 (fp16)

1 740 tokens/s

MLPerf, llama2-70b-99.9 (fp8)

1 057 tokens/s

MLPerf, llama3.1-405b (UNSET)

30.4 tokens/s

MLPerf, llama3.1-405b (fp16)

34.8 tokens/s

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

232.9 tokens/s

MLPerf, mixtral-8x7b (fp8)

5 975 tokens/s

Additional

Slots

OAM Module Dual-slot

Release Date

Dec. 6, 2023 Dec. 18, 2018

Display Outputs

No outputs
1x HDMI 2.0
3x DisplayPort 1.4a
1x USB Type-C

Renting is cheaper than buying