Home › Products › GPU Accelerators › nvidia-h100-80gb-sxm5

NVIDIA H100 80GB SXM5

NVIDIA H100 Tensor Core GPU with 80GB HBM3 memory, SXM5 form factor. 3,958 TFLOPS FP8 (sparse), 900 GB/s NVLink, ideal for large-scale AI training.

Specifications

GPU Architecture

Hopper GH100

Memory

80GB HBM3

Memory Bandwidth

3.35 TB/s

FP8 (sparse)

3,958 TFLOPS

FP16

989 TFLOPS

TF32

495 TFLOPS

FP64

67 TFLOPS

NVLink

900 GB/s, 4th Gen

Interface

PCIe 5.0 / SXM5

TDP

700W

Full Specifications

ParameterValue
GPU ArchitectureHopper GH100
Memory80GB HBM3
Memory Bandwidth3.35 TB/s
FP8 (sparse)3,958 TFLOPS
FP16989 TFLOPS
TF32495 TFLOPS
FP6467 TFLOPS
NVLink900 GB/s, 4th Gen
InterfacePCIe 5.0 / SXM5
TDP700W

Application

Purpose-built for enterprise AI training, large language model (LLM) training, and high-performance computing at scale.

Request a Quote — NVIDIA H100 80GB SXM5

QS Compute — GPU supply partner. Global B2B logistics, competitive pricing, full manufacturer warranty.

Get Your Quote →