AMD Instinct MI300P vs NVIDIA A40 Full Specs

Summary

9,728 Shaders 2.1GHz	10,752 Shaders 1.74GHz
64GB HBM32.66TB/s	48GB GDDR6347.9GB/s
Peak·FP32·FP16 81.72 TFLOPS	Peak·FP32·FP16 37.42 TFLOPS
Form Factor OAM Module	Form Factor PCIe Card
TDP 300W	TDP 300W
Power Connectors -	Power Connectors 1x 8-Pin EPS

Highlights

Peak OPS

Peak OPS FP32 FP16 Tensor FP8-16 FP8-16 Tensor Sparse Tensor FP8-32 FP8-32 Tensor Sparse Tensor FP16-16 FP16-16 Tensor Sparse Tensor FP16-32 FP16-32 Tensor Sparse Tensor FP32 FP64 Tensor FP64 BF16 Tensor BF16 BF16 Tensor Sparse Tensor TF32 Tensor INT8 Tensor INT4 Memory Bandwidth TDP

Instinct MI300P2.62 POPSINT4 Tensor Sparse

x2.18

A401.2 POPSINT4 Tensor Sparse

x1

Show Top Score

Theoretical Performance

Clock Speed2100MHz UC·Base·Boost·OC	Clock Speed1740MHz UC·Base·Boost·OC
Peak OPS 2.62 POPSINT4 Tensor Sparse	Peak OPS 1.2 POPSINT4 Tensor Sparse
Tensor FP8-16 1.31 PFLOPS FP8-16 Tensor Sparse 2.62 PFLOPS Tensor FP8-32 1.31 PFLOPS FP8-32 Tensor Sparse 2.62 PFLOPS	-
Tensor FP16-16 653.7 TFLOPS FP16-16 Tensor Sparse 1.31 PFLOPS Tensor FP16-32 653.7 TFLOPS FP16-32 Tensor Sparse 1.31 PFLOPS	Tensor FP16-16 149.7 TFLOPS FP16-16 Tensor Sparse 299.3 TFLOPS Tensor FP16-32 74.83 TFLOPS FP16-32 Tensor Sparse 149.7 TFLOPS
BF16 - Tensor BF16 653.7 TFLOPS BF16 Tensor Sparse 1.31 PFLOPS	BF16 37.42 TFLOPS Tensor BF16 74.83 TFLOPS BF16 Tensor Sparse 149.7 TFLOPS
Tensor TF32 326.9 TFLOPS	Tensor TF32 37.42 TFLOPS
FP32 81.72 TFLOPS Tensor FP32 81.72 TFLOPS	FP32 37.42 TFLOPS Tensor FP32 -
FP64 40.86 TFLOPS Tensor FP64 81.72 TFLOPS	FP64 584.6 GFLOPS Tensor FP64 -
Tensor INT4 1.31 POPS	Tensor INT4 598.7 TOPS
Tensor INT8 1.31 POPS	Tensor INT8 299.3 TOPS
Ray -	Ray 56.43 TOPS
Pixel Rate -	Pixel Rate 194.9 GPixel/s
Texture Rate 1.28 TTexel/s	Texture Rate 584.6 GTexel/s

Core Config

Shaders 9,728 Shaders	Shaders 10,752 Shaders
TMUs 608 TMUs	TMUs 336 TMUs
ROPs -	ROPs 112 ROPs
Tensor Cores 608 T-Cores	Tensor Cores 336 T-Cores
RT Cores -	RT Cores 84 RT-Cores
CUs 152 CUs	SMs 84 SMs

Clocks

Base Clock 1GHz	Base Clock 1.3GHz
Boost Clock 2.1GHz	Boost Clock 1.74GHz

Cache

L2 Cache 16.4MB shared	L2 Cache 6.1MB shared
L3 Cache 128MB shared	L3 Cache -

Memory

64GB HBM3	48GB GDDR6
Memory Bus 4096-bit	Memory Bus 384-bit
Memory Speed 5.2GT/s	Memory Speed 7.2GT/s
Memory Bandwidth 2.66TB/s	Memory Bandwidth 347.9GB/s
ECC Yes	ECC -

Power & Thermals

TDP 300W

TDP 300W

Video Output

Multi-Monitor -	Multi-Monitor 3
HDCP -	HDCP HDCP 2.3

Ports

-	3x DisplayPort 1.3

Encoder

Encoder Model 2x VCN 2.6

Encoder Model NVENC 7

Decoder

Decoder Model 2x VCN 2.6

Decoder Model NVDEC 5

Card

Form Factor OAM Module	Form Factor PCIe Card
PCIe -	PCIe 2-Slots
-	Height 112 mm (4.41") Width 267 mm (10.51") Depth 40 mm (1.57")
Cooling Passive	Cooling Passive
Power Connectors -	Power Connectors 1x 8-Pin EPS

Chip

Manufacturer AMD	Manufacturer NVIDIA
Chip Designer AMD	Chip Designer NVIDIA
Architecture CDNA 3	Architecture Ampere
Family Instinct	Family Server
Branding	Branding
Codename Aqua Vanjaram	Codename NV172
Chip Variant Aqua Vanjaram XL	Chip Variant GA102-200-KD-A1
Market Segment Server	Market Segment Server
Release Date Dec 6, 2023	Release Date Oct 5, 2020

Fabrication

Foundry TSMC	Foundry Samsung
Other Foundries TSMCActive Interposer	Other Foundries -
Fabrication Node N5 N6Active Interposer	Fabrication Node 8N -
Die Size 740mm²	Die Size 628mm²
Transistor Count 73 Billion	Transistor Count 28.3 Billion
Transistor Density 98.65 MTr/mm²	Transistor Density 45.04 MTr/mm²

Images

No images available

No images available

Competitors

AMDInstinct MI300P vs GPU AMDInstinct MI300A AMDInstinct MI300P vs GPU NVIDIARTX PRO 5000 Blackwell AMDInstinct MI300P vs GPU NVIDIARTX 5880 Ada Show 12 more

NVIDIAA40 vs GPU NVIDIARTX PRO 4000 Blackwell NVIDIAA40 vs GPU NVIDIARTX PRO 4000 Blackwell SFF NVIDIAA40 vs GPU NVIDIAGeForce RTX 5070 Ti Show 12 more