Flopper.io Flopper.io

NVIDIA H100 NVL 94GB vs NVIDIA H20 141GB HBM3e

The H100 delivers 36% more FP32 throughput than the H20 (60 vs 44 TFLOPS dense). The H20 counters with 47 GB more memory (141 GB vs 94 GB).

H100: Hopper, 2023 H20: Hopper, 2025
FP32 dense lead
+36%
H100: 60 vs 44 TFLOPS
Memory
94 vs 141 GB
47 GB more for the H20
Bandwidth
3.9 TB/s vs 4.0 TB/s
H20 moves data faster
TDP
400 W vs 400 W
Same power budget
See rental pricing from $2.27/hr Download comparison sheet

Share this page

NVIDIA H100 NVL 94GB vs NVIDIA H20 141GB HBM3e

A post with this link shows the card on the Image tab as its preview.

Post on X Share on LinkedIn
 Printable version →

NVIDIA H100 NVL 94GB

Leads on FP32 Full specs →

What is the NVIDIA H100?

The H100 is a GPU from NVIDIA on the Hopper architecture, an accelerator for AI training, inference and high-performance computing. It is a PCIe add-in card, so it fits standard servers and workstations. H100 full specifications →

Can I rent the H100 in the cloud?

Yes. Cloud providers tracked by Flopper rent the H100 by the hour, and the pricing page compares their current rates and the price history side by side. H100 rental prices →

How does the H100 compare with other accelerators?

This page sets the H100 against the H20. The H100 is the older generation of the two. Both are NVIDIA parts; the H100 is a PCIe card while the H20 is an SXM module. The compare tool builds the same comparison for any two parts in the catalogue. All GPU comparisons →

Where do the H100 figures come from?

Every H100 figure on this page traces to NVIDIA's own datasheet or specification table. How Flopper verifies specifications →

NVIDIA H20 141GB HBM3e

Full specs →

What is the NVIDIA H20?

The H20 is a GPU from NVIDIA on the Hopper architecture, an accelerator for AI training, inference and high-performance computing. It is an SXM module, the mezzanine format used in dense multi-GPU servers. H20 full specifications →

Can I rent the H20 in the cloud?

Flopper is not tracking any live rental listings for the H20 yet. The provider directory lists the clouds that rent comparable accelerators on demand. GPU cloud providers →

How does the H20 compare with other accelerators?

This page sets the H20 against the H100. The H20 is the newer generation of the two. Both are NVIDIA parts; the H20 is an SXM module while the H100 is a PCIe card. Flopper also compares it head to head with the L600, across every published precision, memory and power. H20 vs L600 →

Where do the H20 figures come from?

NVIDIA published the H20 figures in marketing material rather than a datasheet, and Flopper has no independent confirmation of them. How Flopper verifies specifications →

Specifications

H100
H20
Architecture
HopperHopper
Launch Year
20232025
Form Factor
PCIeSXM
Memory
94 GB141 GB
Memory Bandwidth
3.9 TB/s4.0 TB/s
TDP
400 W400 W
Process Node
4nm4nm

Performance (TFLOPS)

H100
H20
30 TFLOPS 1 TFLOPS
60 TFLOPS 44 TFLOPS
The vendor publishes only the 2:4 sparse figure for this precision
835 TFLOPS sparse
74 TFLOPS
sparse not published
The vendor publishes only the 2:4 sparse figure for this precision
1,671 TFLOPS sparse
148 TFLOPS
sparse not published
The vendor publishes only the 2:4 sparse figure for this precision
1,671 TFLOPS sparse
148 TFLOPS
sparse not published
The vendor publishes only the 2:4 sparse figure for this precision
3,341 TFLOPS sparse
296 TFLOPS
592 TFLOPS sparse
FP6
No verified data available No verified data available
No verified data available No verified data available
The vendor publishes only the 2:4 sparse figure for this precision
3,341 TOPS sparse
296 TOPS
sparse not published
Tensor Core
60 TFLOPS No verified data available

Figures are dense. Where a vendor publishes a figure it sits under the dense one and never decides a winner. Hover any precision for its definition, or read the GPU glossary.

FLOPS by Precision

What actually differs

The H20 is the newer part: Hopper (2025) against the H100's Hopper (2023).

The H20 carries 47 GB more memory (141 GB against 94 GB), so larger models fit on one GPU before you shard. Bandwidth is 3.9 TB/s against 4.0 TB/s, which sets the ceiling on decode speed.

Power is 400 W against 400 W, and at FP32 the H100 does more work per watt. Weight the FP8 and FP16 rows for LLM work and the FP64 row for HPC.

New to the terms? The glossary explains FP64, FP32, TF32, BF16, FP16, FP8, INT8, Sparsity and TFLOPS. For how the generations relate, see NVIDIA architectures explained.

Frequently asked questions

Is the H100 faster than the H20?

At FP32 precision the H100 reaches 60 TFLOPS dense against 44 TFLOPS for the H20. The performance table on this page lists every published precision for both GPUs.

Which should you choose, the H100 or the H20?

Choose the H100 for training and high-throughput inference: it has 36% more FP32 throughput than the H20. Choose the H20 when the model has to fit on one GPU: it carries 47 GB more memory. Packaging differs too: the H100 is a PCIe card while the H20 is an SXM module, so check what your servers accept.

Which has more memory, the H100 or the H20?

The H20 carries 141 GB of memory versus 94 GB for the H100. Memory bandwidth is 3.9 TB/s for the H100 and 4.0 TB/s for the H20.

How much power do the H100 and the H20 draw?

The H100 is rated at 400 W TDP and the H20 at 400 W. On FP32 throughput per watt, the H100 is the more efficient part.

How much do the H100 and the H20 cost to rent?

Live cloud listings tracked by Flopper price the H100 from $2.27 per GPU-hour across 5 providers, and no live listings yet for the H20. The rental section on this page links to every provider, rate and the price history for each GPU.

Where to Rent

NVIDIA H100 NVL 94GB

Rent from

$2.27 / GPU-hr

Lowest of 9 live listings across 5 providers · median $3.17

See all H100 rental prices →

Price history and every provider on the pricing page. Some links are affiliate links.

NVIDIA H20 141GB HBM3e

No live rental listings tracked for the H20 yet.

Browse GPU cloud providers

Get Comparison Updates

New GPUs added weekly. Be the first to see how they compare.

Open in Full Compare Tool → H100 full specs H20 full specs Estimate cluster cost in the calculator →

© 2026 Flopper.io - Compare the GPUs Powering AI