NVIDIA A100X vs NVIDIA H100 SXM Full Specs

Summary

6,912 Shaders 1.44GHz	16,896 Shaders 1.98GHz
80GB HBM2e2.04TB/s	96GB HBM33.35TB/s
Peak·FP32·FP16 19.91 TFLOPS	Peak·FP32·FP16 66.91 TFLOPS
Form Factor PCIe Card	Form Factor SXM5
TDP 300W	TDP 700W
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 FP64 Tensor FP64 BF16 Tensor BF16 BF16 Tensor Sparse Tensor TF32 Tensor INT8 Tensor INT4 Memory Bandwidth TDP

A100X2.55 POPSINT4 Tensor Sparse

x1

H100 SXM3.96 POPSINT8 Tensor Sparse

x1.55

Show Top Score

Theoretical Performance

Clock Speed1440MHz UC·Base·Boost·OC	Clock Speed1980MHz UC·Base·Boost·OC
Peak OPS 2.55 POPSINT4 Tensor Sparse	Peak OPS 3.96 POPSINT8 Tensor Sparse
-	Tensor FP8-16 1.98 PFLOPS FP8-16 Tensor Sparse 3.96 PFLOPS Tensor FP8-32 1.98 PFLOPS FP8-32 Tensor Sparse 3.96 PFLOPS
Tensor FP16-16 318.5 TFLOPS FP16-16 Tensor Sparse 637 TFLOPS Tensor FP16-32 318.5 TFLOPS FP16-32 Tensor Sparse 637 TFLOPS	Tensor FP16-16 989.4 TFLOPS FP16-16 Tensor Sparse 1.98 PFLOPS Tensor FP16-32 989.4 TFLOPS FP16-32 Tensor Sparse 1.98 PFLOPS
BF16 39.81 TFLOPS Tensor BF16 318.5 TFLOPS BF16 Tensor Sparse 637 TFLOPS	BF16 133.8 TFLOPS Tensor BF16 989.4 TFLOPS BF16 Tensor Sparse 1.98 PFLOPS
Tensor TF32 159.3 TFLOPS	Tensor TF32 494.7 TFLOPS
FP32 19.91 TFLOPS	FP32 66.91 TFLOPS
FP64 9.95 TFLOPS Tensor FP64 19.91 TFLOPS	FP64 33.45 TFLOPS Tensor FP64 61.84 TFLOPS
Tensor INT4 1.27 POPS	Tensor INT4 -
Tensor INT8 637 TOPS	Tensor INT8 1.98 POPS
Pixel Rate 230.4 GPixel/s	Pixel Rate 47.5 GPixel/s
Texture Rate 622.1 GTexel/s	Texture Rate 1.04 TTexel/s

Core Config

Shaders 6,912 Shaders	Shaders 16,896 Shaders
TMUs 432 TMUs	TMUs 528 TMUs
ROPs 160 ROPs	ROPs 24 ROPs
Tensor Cores 432 T-Cores	Tensor Cores 528 T-Cores
SMs 108 SMs	SMs 132 SMs

Clocks

Base Clock 795MHz	Base Clock 1.06GHz
Boost Clock 1.44GHz	Boost Clock 1.98GHz
Tensor Clock -	Tensor Clock 1.83GHz

Cache

L2 Cache 41MB shared

L2 Cache 51.2MB shared

Memory

80GB HBM2e	96GB HBM3
Memory Bus 5120-bit	Memory Bus 5120-bit
Memory Speed 3.2GT/s	Memory Speed 5.2GT/s
Memory Bandwidth 2.04TB/s	Memory Bandwidth 3.35TB/s

Power & Thermals

TDP 300W

TDP 700W

Decoder

Decoder Model 5x NVDEC 4

Decoder Model 7x NVDEC 5

Card

Form Factor PCIe Card	Form Factor SXM5
PCIe 2-Slots	PCIe -
Height 111 mm (4.37") 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 NVIDIA
Chip Designer NVIDIA	Chip Designer NVIDIA
Architecture Ampere	Architecture Hopper
Family Server	Family Server
Branding	Branding
Codename NV170	Codename NV180
Market Segment Server	Market Segment Server
Release Date Jun 28, 2021	Release Date Mar 22, 2022

Fabrication

Foundry TSMC	Foundry TSMC
Fabrication Node 7N	Fabrication Node 4N
Die Size 826mm²	Die Size 814mm²
Transistor Count 54.2 Billion	Transistor Count 80 Billion
Transistor Density 65.62 MTr/mm²	Transistor Density 98.28 MTr/mm²

Images

No images available

No images available

Competitors

NVIDIAA100X vs GPU AMDInstinct MI300P NVIDIAA100X vs GPU NVIDIARTX PRO 6000 Blackwell Max-Q NVIDIAA100X vs GPU NVIDIAGeForce RTX 5070 Show 12 more

NVIDIAH100 SXM vs GPU NVIDIAB100 NVIDIAH100 SXM vs GPU NVIDIAB200 NVIDIAH100 SXM vs GPU NVIDIAGB200 Show 12 more