AMD Radeon Instinct MI300X vs NVIDIA Tesla K40m

Comparison of AMD Radeon Instinct MI300X with 192 GB HBM3 and 19,456 cores vs NVIDIA Tesla K40m with 12 GB GDDR5 and 2,880 cores.

Loading...

Performance Rating

H200 H200
MI325X MI325X
A100 A100

AMD Radeon Instinct MI300X

94.1

AMD Radeon Instinct MI300X

94.1
MI250 MI250
Instinct MI300X Instinct MI300X
RX 7900 XTX RX 7900 XTX

NVIDIA Tesla K40m

NVIDIA Tesla K40m

Contents:

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

Memory

Memory Size

🔥 +1,500% 192 ГБ
12 GB

Memory Type

HBM3 GDDR5

Memory Bandwidth

🔥 10.3 TB/s
288.4 GB/s

Memory Bus Width

8,192 бит 384 бит

ML Performance

FP16 (Half Precision)

🔥 653.7 TFLOPS
No

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

🔥 +1,520% 81.72 TFLOPS
5.046 TFLOPS

FP64 (Double Precision)

🔥 +4,759% 81.72 TFLOPS
1.682 TFLOPS

CUDA Cores

🔥 +576% 19,456
2,880

RT Cores

No No

Architecture & Compatibility

GPU Architecture

CDNA 3.0 Kepler

SM (Streaming Multiprocessor)

No No

PCIe Version

PCIe 5.0 x16 PCIe 3.0 x16

ML Software Support

CUDA Version

No 3.5

Clocks & Performance

Base Clock

🔥 +34% 1,000
745

Boost Clock

🔥 +140% 2,100
876

Memory Clock

🔥 +68% 2,525
1,502

Power Consumption

TDP/TGP

750 W
🔥 -67% 245 W

Recommended PSU

1150 W
🔥 -52% 550 W

Power Connector

None No

Rendering

Texture Units (TMU)

🔥 +407% 1,216
240

ROP

No No

L2 Cache

🔥 16 MB
1536 KB

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 Nov. 22, 2013

Display Outputs

No outputs
No outputs

Renting is cheaper than buying