A10 vs A30
A10 starts at $1.00/GPU-hour on-demand; A30 starts at $0.35/GPU-hour on-demand. Compare the exact variants, memory and published specifications below. Peak compute figures do not establish application speed.
A10
A30
Spec sheet
| A10 | A30 | |
|---|---|---|
| Spec variant | A10 | A30 |
| Architecture | Ampere | Ampere |
| VRAM (spec variant) | 24 GB | 24 GB |
| FP16 TFLOPS (dense) | 125 | 165 |
| Memory bandwidth | 600 GB/s | 933 GB/s |
| TDP | 150 W | 165 W |
| Interconnect | PCIe Gen4 | NVLink 200GB/s |
| Launched | 2021 | 2021 |
Rental price per manufacturer-rated FP16 compute: A10 $8.00 per PFLOP-hour vs A30 $2.12. Formula: cheapest $/hr ÷ dense FP16 PFLOPS (1 PFLOP = 1000 TFLOPS). This uses A10 and A30, whose datasheets may differ from the variants above. It is not measured job throughput. Specs: A10 manufacturer, A30 manufacturer.
What they cost to rent (2026-10-02)
| A10 | A30 | |
|---|---|---|
| Cheapest on-demand $/hr | $1.00 | $0.35 |
| Cheapest spot/community $/hr | – | – |
| Cheapest source | OVHcloud | Massed Compute |
| Providers offering it | 5 | 1 |
Lowest recorded on-demand rate per configuration; highlighted = cheaper. Spot/marketplace tiers shown on the individual pages.
FAQ
A30 is cheaper: the A10 rents from $1.00 per GPU-hour and the A30 from $0.35 on-demand, updated 2026-10-02.
The largest recorded A10 variant has 24 GB and the largest A30 variant has 24 GB. Other variants may have less memory. They match on maximum VRAM.
Compare more rental options: the screener · price movers · methodology.