Skip to main content

RTX4090 vs RTX5090

The RTX4090 delivers 1.3× the FP16 compute of the RTX4090 while costing 2.2× less on average, making it 74% more cost-efficient per TFLOP.

Spec Comparison

Spec
RTX4090
RTX5090
Architecture
Ada Lovelace
Blackwell
VRAM
24 GB
32 GB
FP16 TFLOPS
165
210
Memory Bandwidth
1,008 GB/s
1,792 GB/s
MSRP
$1,599
$1,999
Used/Market Price
$1,200
N/A

Live Pricing

RTX4090
Cheapest
$0.131/hr
Median
$0.718/hr
Average
$1.01/hr
Providers
3
Instances
96
RTX5090
Cheapest
$0.0940/hr
Median
$1.59/hr
Average
$2.19/hr
Providers
3
Instances
108

Cost Efficiency

Metric
RTX4090
RTX5090
$/TFLOP/hr
$0.0044
$0.0076
$/GB-VRAM/hr
$0.0299
$0.0497

Lower $/TFLOP and $/GB-VRAM is better. Based on median hourly pricing.

Provider Availability

Both available on (3)

RTX4090 Cloud Pricing

96 instances
ProviderGPUsVRAMRegionTypePrice/hr
CHEAPESTVast.ai1x24 GBSpot$0.131Rent
RunPod1x24 GBCASpot$0.200Rent
RunPod1x24 GBFRSpot$0.200Rent
RunPod1x24 GBSESpot$0.200Rent
RunPod1x24 GBSKSpot$0.200Rent
RunPod1x24 GBTTSpot$0.200Rent
RunPod1x24 GBUSSpot$0.200Rent
RunPod1x24 GBEU-RO-1Spot$0.290Rent
RunPod1x24 GBEUR-IS-2Spot$0.290Rent
RunPod1x24 GBUS-CA-2Spot$0.290Rent

RTX5090 Cloud Pricing

108 instances
ProviderGPUsVRAMRegionTypePrice/hr
CHEAPESTVast.ai1x32 GBUS-EastSpot$0.0940Rent
Vast.ai1x32 GBAsia-PacificSpot$0.100Rent
Vast.ai1x32 GBEU-CentralSpot$0.120Rent
Vast.ai2x64 GBAsia-PacificSpot$0.267Rent
Vast.ai1x32 GBUS-CentralSpot$0.268Rent
Vast.ai1x32 GBUS-CentralOn-Demand$0.280Rent
Vast.ai1x32 GBUS-WestSpot$0.293Rent
Vast.ai1x32 GBEU-CentralOn-Demand$0.330Rent
Vast.ai1x32 GBAsia-PacificOn-Demand$0.368Rent
Vast.ai2x64 GBvn-vietnamSpot$0.373Rent

Compare with Other GPUs

RTX4090 vs ...

RTX5090 vs ...