NVIDIA H100 PCIe 80GB vs NVIDIA L4 24GB
H100 PCIe 80GB has 233% more memory than L4 24GB (80 GB vs 24 GB). Its memory bandwidth is 567% more (2 vs 0.3 TB/s). In the cloud, the median on-demand H100 PCIe 80GB price is $3.04 per GPU-hour against $0.44 for the L4 24GB. Per gigabyte of GPU memory, that is $0.038 vs $0.018 per hour.
Specifications
Manufacturer figures as recorded by Kovara. Higher memory and bandwidth are highlighted; neither is a benchmark of your workload.
| H100 PCIe 80GB | L4 24GB | |
|---|---|---|
| Manufacturer | NVIDIA | NVIDIA |
| Architecture | Hopper | Ada Lovelace |
| Memory | 80 GB HBM2E | 24 GB GDDR6 |
| Memory bandwidth | 2 TB/s | 0.3 TB/s |
| Power (TDP) | 350 W | 72 W |
| Form factor | PCIe dual-slot | PCIe low-profile single-slot |
Cloud prices today
Public offers Kovara tracks, normalised to US dollars per GPU-hour. Prices use on-demand terms where available; spot and reserved offers are on each GPU’s price pages. Lower is highlighted.
| H100 PCIe 80GB | L4 24GB | |
|---|---|---|
| Lowest price | $2.49 | $0.44 |
| Median price | $3.04 | $0.44 |
| Highest price | $3.29 | $0.57 |
| Median per GB of memory | $0.038 | $0.018 |
| Public offers (all terms) | 6 | 7 |
| Providers | 5 | 3 |
Where to rent each
Lowest tracked price per provider. Open a provider to see every offer, region and term.
Questions
Which is cheaper to rent, the H100 PCIe 80GB or the L4 24GB?
Right now the cheapest on-demand H100 PCIe 80GB Kovara tracks is $2.49 per GPU-hour and the cheapest L4 24GB is $0.44. Medians are $3.04 and $0.44. The cheaper GPU per hour is not always cheaper per job: a faster or larger GPU can finish sooner or need fewer units.
Which has more memory?
The H100 PCIe 80GB has more memory: 80 GB against 24 GB. Memory decides which models and batch sizes fit on one GPU.
Are these benchmark results?
No. This page compares manufacturer specifications and public prices. Real performance depends on the model, precision, software and interconnect. Kova can help you estimate cost for a specific workload.