RTX5090 vs T4
The RTX5090 delivers 3.2× the FP16 compute of the T4 while costing 1.9× more on average, making it 70% more cost-efficient per TFLOP.
Spec Comparison
Spec
RTX5090
T4
Architecture
Blackwell
Older
VRAM
32 GB
16 GB
FP16 TFLOPS
210
65
Memory Bandwidth
1,792 GB/s
320 GB/s
MSRP
$1,999
N/A
Used/Market Price
N/A
$1,500
Live Pricing
Cost Efficiency
Metric
RTX5090
T4
$/TFLOP/hr
$0.0082
$0.0140
$/GB-VRAM/hr
$0.0540
$0.0568
Lower $/TFLOP and $/GB-VRAM is better. Based on median hourly pricing.
Provider Availability
RTX5090 Cloud Pricing
108 instances| Provider | GPUs | VRAM | Region | Type | Price/hr | |
|---|---|---|---|---|---|---|
| CHEAPESTVast.ai | 1x | 32 GB | China | Spot | $0.213 | Rent |
| Vast.ai | 1x | 32 GB | China | On-Demand | $0.236 | Rent |
| Vast.ai | 1x | 32 GB | EU-Central | Spot | $0.267 | Rent |
| Vast.ai | 1x | 32 GB | US-Central | Spot | $0.268 | Rent |
| Vast.ai | 1x | 32 GB | US-Central | On-Demand | $0.278 | Rent |
| Vast.ai | 1x | 32 GB | China | Spot | $0.289 | Rent |
| Vast.ai | 1x | 32 GB | EU-Central | Spot | $0.333 | Rent |
| Vast.ai | 1x | 32 GB | China | On-Demand | $0.401 | Rent |
| Vast.ai | 2x | 64 GB | mx-mexico | Spot | $0.453 | Rent |
| RunPod | 1x | 32 GB | EU-RO-1 | Spot | $0.530 | Rent |
T4 Cloud Pricing
1557 instances| Provider | GPUs | VRAM | Region | Type | Price/hr | |
|---|---|---|---|---|---|---|
| CHEAPESTAzure | 1x | 16 GB | Asia-Pacific | Spot | $0.0684 | Rent |
| Azure | 1x | 16 GB | US-Central | Spot | $0.0686 | Rent |
| Azure | 1x | 16 GB | Asia-Pacific | Spot | $0.0977 | Rent |
| Azure | 1x | 16 GB | US-Central | Spot | $0.0980 | Rent |
| GCP | 1x | 16 GB | me-west1-b | Spot | $0.112 | Rent |
| GCP | 1x | 16 GB | me-west1-c | Spot | $0.112 | Rent |
| GCP | 1x | 16 GB | EU-West | Spot | $0.117 | Rent |
| GCP | 1x | 16 GB | me-west1-b | Spot | $0.118 | Rent |
| GCP | 1x | 16 GB | me-west1-c | Spot | $0.118 | Rent |
| AWS | 1x | 16 GB | South America | Spot | $0.123 | Rent |