NVIDIA H800 PCIe 80GB vs NVIDIA H100 NVL 94GB

The H100 delivers 18% more FP32 throughput than the H800 (60 vs 51 TFLOPS dense). The H100 also carries 14 GB more memory (94 GB vs 80 GB).

H800: Hopper, 2022 H100: Hopper, 2023
FP32 dense lead
+18%
H100: 60 vs 51 TFLOPS
VRAM
80 vs 94 GB
14 GB more for the H100
Bandwidth
2.0 TB/s vs 3.9 TB/s
H100 moves data faster
TDP
350 W vs 400 W
H800 draws 50 W less

Specifications

H800
H100
Architecture
HopperHopper
Launch Year
20222023
Form Factor
PCIePCIe
VRAM
80 GB94 GB
Memory Bandwidth
2.0 TB/s3.9 TB/s
TDP
350 W400 W
Process Node
4nm4nm

Performance (TFLOPS)

H800
H100
FP64
0.8 TFlops 30.0 TFlops
FP32
51.0 TFlops 60.0 TFlops
FP16
756.5 TFlops
sparse not published
BF16
756.5 TFlops
sparse not published
FP8
1513.0 TFlops
sparse not published
INT8
1513.0 TFlops
sparse not published
Tensor Core
FP64 (TC)
60.0 TFlops

FLOPS by Precision

What actually differs

The H100 is the newer part: Hopper, launched in 2023, against the H800's Hopper from 2022. Newer architectures typically add lower-precision formats and better throughput per watt, so check the precision rows your workload actually uses.

Dense throughput for the H800 against the H100: FP64 0.8 vs 30 TFLOPS, FP32 51 vs 60 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 80 GB against 94 GB, fed at 2.0 TB/s versus 3.9 TB/s. More memory per GPU means larger models fit before you have to shard across cards, which often matters more than raw TFLOPS for inference.

Power budgets are 350 W for the H800 and 400 W for the H100. At FP32 that works out to 0.15 against 0.15 TFLOPS per watt.

For LLM training and inference, weight the FP8 and FP16 rows and memory capacity most heavily. For scientific and HPC workloads, the FP64 row is the one to read.

Frequently asked questions

Is the H800 faster than the H100?

At FP32 precision the H100 reaches 60 TFLOPS dense against 51 TFLOPS for the H800. The performance table on this page lists every published precision for both GPUs.

Which has more memory, the H800 or the H100?

The H100 carries 94 GB of VRAM versus 80 GB for the H800. Memory bandwidth is 2.0 TB/s for the H800 and 3.9 TB/s for the H100.

How much power do the H800 and the H100 draw?

The H800 is rated at 350 W TDP and the H100 at 400 W. On FP32 throughput per watt, the H100 is the more efficient part.

Can I rent the H800 or the H100 in the cloud?

Yes. Live cloud listings tracked by Flopper start at $2.59 per GPU hour. The rental pricing section on this page lists current providers and rates for both GPUs.

Where to Rent

NVIDIA H100 NVL 94GB

ProviderConfigurationPrice/GPU-hrChecked
RunPod
1× H100 NVL Community $2.59 2h ago View →
RunPod
1× H100 NVL $3.19 2h ago View →
Microsoft Azure
1× H100 NVL $6.98 2h ago View →
Microsoft Azure
2× H100 NVL $6.98 2h ago View →
Microsoft Azure
1× H100 NVL $6.98 2h ago View →
All H100 listings and price history →

Get Comparison Updates

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