Back to all GPUs

NVIDIA RTX 5090 Cloud GPU Specs

HexGrid Cloud rents the NVIDIA GeForce RTX 5090 at $0.55 per GPU hour, billed per minute — about 44% below the $0.99/hr average across 1 other clouds. The RTX 5090 carries 32GB of GDDR7 memory on the Blackwell architecture.

32GB VRAMBlackwellDeploys in ~1 minSecure Cloud

HexGrid Cloud price

$0.55/hr

Billed per minute

VRAM

32GB

GDDR7

FP32

104.8 TFLOPS

Compute performance

Market average

$0.99/hr

You save 44%

Running the RTX 5090 on HexGrid Cloud

What you get when you deploy with us, beyond the hourly rate.

Live in under a minute

Pick a region, hit deploy, and SSH in. No quota requests, no sales call.

Per-minute billing

You pay for the minutes you use. Stop the instance and billing stops with it.

Ready for training day one

CUDA, PyTorch and the usual drivers are preinstalled, or bring your own image.

No egress charges

Move checkpoints and datasets out without a surprise bandwidth bill.

RTX 5090 pricing across the market

Published on-demand rates from other clouds, so you can see where our RTX 5090 price sits.

Market average $0.99/hr across 1 cloud

44% below market average
  • HexGrid CloudYou're here$0.55/hr
  • RunPod$0.99/hr

Other providers' rates are their published on-demand list prices. They exclude committed-use discounts and can change at any time. Shown for reference only.

About the NVIDIA GeForce RTX 5090

What the RTX 5090 is built for, and where it falls short.

The NVIDIA GeForce RTX 5090 is a Blackwell GPU with 32GB of GDDR7 memory, a 512-bit memory interface, fifth-generation Tensor Cores, and fourth-generation RT Cores. Its combination of AI acceleration and 1.79 TB/s memory bandwidth makes it particularly capable for generative AI, inference, video generation, rendering, and development.

Best suited for

The RTX 5090 combines 32GB of GDDR7 memory with Blackwell Tensor Cores and very high memory bandwidth, making it one of the fastest consumer GPUs for local and cloud AI workloads.

  • LLM inference
  • AI fine-tuning
  • Image generation
  • Video generation
  • 3D rendering
  • AI development

Strengths

  • 32GB GDDR7 memory
  • Very high memory bandwidth
  • Blackwell Tensor Cores
  • FP4 and FP8 support
  • Three NVENC engines

Limitations

  • No ECC memory
  • No NVLink
  • No MIG support
  • High 575W power consumption

NVIDIA GeForce RTX 5090 specifications

Full technical specifications for the NVIDIA GeForce RTX 5090.

VendorNVIDIAArchitectureBlackwellChipGB202CategoryConsumer GPUReleased2025Form factor2-slotInterfacePCIe 5.0 x16

Memory

Memory
32 GB
Type
GDDR7
Bandwidth
1,792 GB/s
Bus width
512-bit
ECC
No

Compute

CUDA cores
21,760
Tensor cores
680
Tensor generation
Gen 5
RT cores
170
Boost clock
2,407 MHz

AI & compute performance

FP32
104.8 TFLOPS
RT
318 TFLOPS

Power & physical

TDP
575 W
Power connector
1x 600W PCIe Gen 5 cable
Form factor
2-slot
Cooling
Active
Length
304 mm

Interconnect

Interface
PCIe 5.0 x16
NVLink
No
Infinity Fabric
No

Virtualization

MIG
No
vGPU
No
SR-IOV
No

Media engines

NVENC engines
3
NVDEC engines
2
AV1 encode
Yes
AV1 decode
Yes

Software

CUDA
Yes
ROCm
No

Precision support

FP64
No
FP32
Yes
TF32
Yes
BF16
Yes
FP16
Yes
FP8
Yes
FP6
Yes
FP4
Yes
INT8
Yes

RTX 5090 frequently asked questions

Common questions about NVIDIA GeForce RTX 5090 pricing, specifications and workloads on HexGrid Cloud.

How much VRAM does the RTX 5090 have?

The RTX 5090 has 32GB of GDDR7 memory on a 512-bit memory interface.

Is the RTX 5090 good for AI?

Yes. Its Blackwell Tensor Cores, 32GB memory capacity, and high memory bandwidth make it highly capable for inference, generative AI, and development.

Does the RTX 5090 support FP4?

Yes. NVIDIA Blackwell introduces low-precision formats including FP4 for supported AI workloads.

Does the RTX 5090 support NVLink?

No. The GeForce RTX 5090 does not support NVLink.

How much does an RTX 5090 cost per hour?

Cloud RTX 5090 pricing varies by provider, region, and configuration. HexGrid compares available hourly GPU offers.

Other GPUs on HexGrid Cloud

If the RTX 5090 isn't the right fit, these are also available to deploy.

Browse every GPU we rent

Start training on the RTX 5090 today

$0.55 per GPU hour, billed per minute, running in about a minute.

Specification sources