NVIDIA · Data centre
T4
Rented by one provider in this index, Azure. A single price is a price, not a comparison.
Best published rate
$0.343
per GPU / hour
Azure · 16 GB
Azure · 16 GB
Every published rate
cheapest first · USD per GPU per hour| Provider | What they call it | Their own price, and the arithmetic | Per GPU-hour |
|---|---|---|---|
| Azure
16 GB |
T4
Standard_NC4as_T4_v3 |
$0.3430 per hour = $0.3430 per GPU-hour
region ATT Dallas 1 · 776 offers spot $0.068 |
$0.3430
$5.28 per PFLOP-hr |
The silicon
vendor-published| Variant | Architecture | Memory | FP16 dense | Board power |
|---|---|---|---|---|
| T4 | Turing 2018 | 16 GB GDDR6 | 65 TFLOPS | 70 W |
Vendor-published figures. FP16 throughput is dense tensor throughput with no sparsity assumed; vendors often headline the sparse figure, which is twice as large. Vendor page ↗
What they call it
Providers write the same hardware down differently. Every name below is one this index has seen for this chip; nothing was renamed to make the table tidier.
T4