NVIDIA A40 vs AMD Instinct MI300P Full Specs

Summary

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

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

A401.2 POPSINT4 Tensor Sparse

x1

Instinct MI300P2.62 POPSINT4 Tensor Sparse

x2.18

Show Top Score

Theoretical Performance

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

Core Config

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

Clocks

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

Cache

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

Memory

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

Power & Thermals

TDP 300W

TDP 300W

Video Output

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

Ports

3x DisplayPort 1.3

-

Encoder

Encoder Model NVENC 7

Encoder Model 2x VCN 2.6

Decoder

Decoder Model NVDEC 5

Decoder Model 2x VCN 2.6

Card

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

Chip

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

Fabrication

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

Images

No images available

No images available

Competitors

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

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