A100 vs H200
A100 starts at $0.89/GPU-hour on-demand; H200 starts at $2.60/GPU-hour on-demand. Compare the exact variants, memory and published specifications below. Peak compute figures do not establish application speed.
65.8% cheaper at A100.
updated 2026-10-06. Price alone does not establish equivalent capacity or performance.
A100
H200
Spec sheet
| A100 | H200 | |
|---|---|---|
| Spec variant | A100 SXM 40GB | H200 (SXM) |
| Architecture | Ampere | Hopper |
| VRAM (spec variant) | 40 GB | 141 GB |
| FP16 TFLOPS (dense) | 312 | 989 |
| Memory bandwidth | 1555 GB/s | 4800 GB/s |
| TDP | 400 W | 700 W |
| Interconnect | NVLink 600GB/s | NVLink 900GB/s |
| Launched | 2020 | 2023 |
Rental price per manufacturer-rated FP16 compute: A100 $2.85 per PFLOP-hour vs H200 $2.63. Formula: cheapest $/hr ÷ dense FP16 PFLOPS (1 PFLOP = 1000 TFLOPS). This uses A100 SXM 40GB and H200 (SXM), whose datasheets may differ from the variants above. It is not measured job throughput. Specs: A100 manufacturer, H200 manufacturer.
What they cost to rent (2026-10-06)
| A100 | H200 | |
|---|---|---|
| Cheapest on-demand $/hr | $0.89 | $2.60 |
| Cheapest spot/community $/hr | $0.679 | $2.61 |
| Cheapest source | Jarvislabs | GMI Cloud |
| Providers offering it | 22 | 25 |
Lowest recorded on-demand rate per configuration; highlighted = cheaper. Spot/marketplace tiers shown on the individual pages.
FAQ
A100 is cheaper: the A100 rents from $0.89 per GPU-hour and the H200 from $2.60 on-demand, updated 2026-10-06.
The largest recorded A100 variant has 80 GB and the largest H200 variant has 143 GB. Other variants may have less memory. The H200 has more.
Compare more rental options: the screener · price movers · methodology.