NVIDIA H100 vs Qualcomm Adreno 642L Full Specs

Summary

14,592 Shaders 1.75GHz	512 Shaders 550MHz
80GB HBM2e2.04TB/s placeholder	Shared Memory25.6GB/sLPDDR5
Peak·FP32·FP16 51.22 TFLOPS	Peak·FP32·FP16 563.2 GFLOPS
Form Factor PCIe Card	Form Factor iGPU
TDP 350W	TDP 5W
Power Connectors 1x 16-Pin 12VHPWR	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 Memory Bandwidth TDP

H1003.03 POPSINT8 Tensor Sparse

x2686.25

Adreno 642L1.13 TFLOPSFP16

x1

Show Top Score

Theoretical Performance

Clock Speed1755MHz UC·Base·Boost·OC	Clock Speed550MHz UC·Base·Boost·OC
Peak OPS 3.03 POPSINT8 Tensor Sparse	Peak OPS 1.13 TFLOPSFP16
Tensor FP8-16 1.51 PFLOPS FP8-16 Tensor Sparse 3.03 PFLOPS Tensor FP8-32 1.51 PFLOPS FP8-32 Tensor Sparse 3.03 PFLOPS	-
Tensor FP16-16 756.4 TFLOPS FP16-16 Tensor Sparse 1.51 PFLOPS Tensor FP16-32 756.4 TFLOPS FP16-32 Tensor Sparse 1.51 PFLOPS	-
BF16 102.4 TFLOPS Tensor BF16 756.4 TFLOPS BF16 Tensor Sparse 1.51 PFLOPS	-
Tensor TF32 378.2 TFLOPS	Tensor TF32 -
FP32 51.22 TFLOPS	FP32 563.2 GFLOPS
FP64 25.61 TFLOPS Tensor FP64 47.28 TFLOPS	FP64 140.8 GFLOPS Tensor FP64 -
Tensor INT8 1.51 POPS	-
Pixel Rate 42.1 GPixel/s	Pixel Rate 4.4 GPixel/s
Texture Rate 800.3 GTexel/s	Texture Rate 8.8 GTexel/s

Core Config

Shaders 14,592 Shaders	Shaders 512 Shaders
TMUs 456 TMUs	TMUs 16 TMUs
ROPs 24 ROPs	ROPs 8 ROPs
Tensor Cores 456 T-Cores	Tensor Cores -
SMs 114 SMs	EUs 4 EUs

Clocks

Base Clock 1.06GHz	Base Clock 490MHz
Boost Clock 1.75GHz	Boost Clock 550MHz
Tensor Clock 1.62GHz	Tensor Clock -

Cache

L1 64KB/SM Tex	L1 8KB/CU
L2 Cache 50MB shared	L2 Cache 128KB shared
L3 Cache -	L3 Cache 1MB shared

Memory

80GB HBM2e placeholder	Shared MemoryLPDDR5
Memory Bus 5120-bit	Memory Bus 32-bit
Memory Speed 3.2GT/s	Memory Speed 6.4GT/s
Memory Bandwidth 2.04TB/s	Memory Bandwidth 25.6GB/s
ECC No	ECC No

Power & Thermals

TDP 350W

TDP 5W

Video Output

Max Resolution -	Max Resolution 2520x1080
Refresh Rate Calculator -	Refresh Rate Calculator 1920x1080 Resolution1080p FHD Refresh Rate 78Hz
Multi-Monitor -	Multi-Monitor 1
Adaptive Sync G-Sync FreeSync	Adaptive Sync -
DSC Not Supported	DSC Not Supported

API Support

-	DirectX DirectX 12 Direct3D 12_1
OpenGL - OpenCL 3 Vulkan -	OpenGL 3.2 OpenCL 2 Vulkan 1.1
CUDA 9 PureVideo HD VP11 VDPAU Feature Set K	-

Encoder

-	Encoder Hexagon 770
Codec -	Codec H.263 MPEG-4 VC-1 VP8 VP9 AVC (H.264) HEVC (H.265)

Decoder

Decoder 7x NVDEC 5	Decoder Hexagon 770
Codec MPEG-1 MPEG-2 MPEG-4 VC-1 VP8 VP9 AVC (H.264) HEVC (H.265) AV1	Codec MPEG-1 MPEG-2 MPEG-4 VC-1 VP8 VP9 AVC (H.264) HEVC (H.265) -

Card

Form Factor PCIe Card	Form Factor iGPU
PCIe PCIe 5.0 x16 2-Slots	PCIe -
Height 111mm (4.37") Width 268mm (10.55") Depth 40mm (1.57")	-
Cooling Passive	Cooling Open-Air
Power Connectors 1x 16-Pin 12VHPWR	Power Connectors -
Multi-GPU Multi-GPU Support Supported Multi-GPU Type NVLink	Multi-GPU -

Chip

Manufacturer NVIDIA	Manufacturer Qualcomm
Chip Designer NVIDIA	Chip Designer Qualcomm
Architecture Hopper	Architecture Adreno 600
Family Server	Family Adreno 600
Branding	Branding
Codename NV180	Codename -
Market Segment Server	Market Segment Smartphone
Release Date Mar 22, 2022	Release Date May 19, 2021

Fabrication

Foundry TSMC	Foundry TSMC
Fabrication Node 4N	Fabrication Node N6
Die Size 814mm²	Die Size -
Transistor Count 80B	Transistor Count -
Transistor Density 98.28 MTr/mm²	Transistor Density -

Images