How does NVIDIA H100 SXM5 80GB compare to NVIDIA H100 PCIe 80GB in memory bandwidth?

The NVIDIA H100 SXM5 80GB has 3,350 GB/s memory bandwidth, while the NVIDIA H100 PCIe 80GB has 2,000 GB/s. Higher bandwidth benefits memory-bound workloads like large-batch LLM inference.

Which is more power-efficient, NVIDIA H100 SXM5 80GB or NVIDIA H100 PCIe 80GB?

Compare NVIDIA H100 SXM5 80GB (989 TFLOPS FP16, 700 W TDP) vs NVIDIA H100 PCIe 80GB (756 TFLOPS FP16, 350 W TDP) for power efficiency. Lower TDP with higher TFLOPS means better TFLOPS/watt.

All comparisons →

Hopper

NVIDIA H100 SXM5 80GB

wins

The H100 SXM5 is the high-performance form factor of the H100.

LLM trainingMulti-GPU NVLink clustersHigh-throughput inference

Hopper

NVIDIA H100 PCIe 80GB

wins

The H100 PCIe is the more affordable form factor of the H100.

Budget Hopper inferencePCIe server deploymentsCost-sensitive LLM serving

Performance comparison

Visual bars — winner highlighted

NVIDIA H100 SXM5 80GB

Metric

NVIDIA H100 PCIe 80GB

FP16 TFLOPS

989 T

756 T

Bandwidth

3.4k GB/s

2.0k GB/s

VRAM

803 GB

802 GB

TDP (lower=better)

700 W

350 W

TFLOPS/watt

1.41

2.16

GB/s per watt

4.79

5.71

Full specification table

Spec	NVIDIA H100 SXM5 80GB	NVIDIA H100 PCIe 80GB
Architecture	Hopper	Hopper
VRAM	80 GB HBM3	80 GB HBM2e
VRAM type	HBM3	HBM2e
Memory bandwidth	3,350 GB/s	2,000 GB/s
FP16 TFLOPS	989 TFLOPS	756 TFLOPS
TDP	700 W	350 W
TFLOPS/watt	1.41 T/W	2.16 T/W
GB/s per watt	4.79 GB/s/W	5.71 GB/s/W
NVLink	Yes	No
Release year	2022	2022

Live cloud pricing

On-demand $/hr across providers — updated in real time

Loading prices…

Power efficiency analysis

TFLOPS/watt and GB/s/watt — critical for data center TCO

NVIDIA H100 SXM5 80GB

FP16 TFLOPS989 TFLOPS

TDP700 W

TFLOPS/watt1.413 T/W

GB/s per watt4.79 GB/s/W

NVIDIA H100 PCIe 80GB

FP16 TFLOPS756 TFLOPS

TDP350 W

TFLOPS/watt2.160 T/W

GB/s per watt5.71 GB/s/W

TFLOPS/watt measures compute efficiency — how much AI throughput you get per watt of power consumed. For data centers with PUE of 1.2–1.5, a 10% improvement in TFLOPS/watt translates directly to lower electricity costs and cooling requirements. GB/s/watt measures memory bandwidth efficiency, which is the binding constraint for memory-bound LLM inference workloads.

When to choose each GPU

Choose NVIDIA H100 SXM5 80GB for:

LLM training
Multi-GPU NVLink clusters
High-throughput inference

Choose NVIDIA H100 PCIe 80GB for:

Budget Hopper inference
PCIe server deployments
Cost-sensitive LLM serving

Popular GPU comparisons

H100 80GB vs A100 80GB H100 80GB vs H200 141GB H100 80GB vs B200 180GB H100 80GB vs L40S H200 141GB vs B200 180GB MI300X vs H100 80GB MI300X vs H200 141GB L40S vs A100 80GB

GPU Leaderboard

TFLOPS/$, GB/s/$, TFLOPS/watt

Compare other GPUs

Pick any two GPUs

Cost calculator

Model total rental cost