Back to all GPUs

NVIDIA RTX 6000 Ada Cloud GPU Specs

HexGrid Cloud rents the NVIDIA RTX 6000 Ada Generation at $0.88 per GPU hour, billed per minute. The RTX 6000 Ada carries 48GB of GDDR6 memory on the Ada Lovelace architecture.

48GB VRAMAda LovelaceDeploys in ~1 minSecure Cloud

HexGrid Cloud price

$0.88/hr

Billed per minute

VRAM

48GB

GDDR6

FP32

91.1 TFLOPS

Compute performance

Market average

$0.84/hr

Other clouds, list price

Running the RTX 6000 Ada 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 6000 Ada pricing across the market

Published on-demand rates from other clouds, so you can see where our RTX 6000 Ada price sits.

Market average $0.84/hr across 1 cloud

  • HexGrid CloudYou're here$0.88/hr
  • RunPod$0.84/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 RTX 6000 Ada Generation

What the RTX 6000 Ada is built for, and where it falls short.

The NVIDIA RTX 6000 Ada Generation is a 48GB professional GPU based on Ada Lovelace. It delivers high CUDA, Tensor, ray-tracing, encoding, and visualization performance while supporting ECC memory and NVIDIA virtual GPU software.

Best suited for

The RTX 6000 Ada combines 48GB ECC memory with powerful Ada Tensor and RT Cores, making it strong for AI workloads that also need professional graphics and virtualization.

  • LLM inference
  • LLM fine-tuning
  • Image generation
  • Video generation
  • 3D rendering
  • Virtual workstations

Strengths

  • 48GB ECC memory
  • High FP32 performance
  • FP8-capable Tensor Cores
  • Excellent rendering performance
  • vGPU support

Limitations

  • No NVLink
  • No MIG
  • Lower large-scale AI training performance than Hopper GPUs

NVIDIA RTX 6000 Ada Generation specifications

Full technical specifications for the NVIDIA RTX 6000 Ada Generation.

VendorNVIDIAArchitectureAda LovelaceChipAD102CategoryWorkstation GPUReleased2022Form factorDual-slotInterfacePCIe 4.0 x16

Memory

Memory
48 GB
Type
GDDR6
Bandwidth
960 GB/s
Bus width
384-bit
ECC
Yes

Compute

CUDA cores
18,176
Tensor cores
568
Tensor generation
Gen 4
RT cores
142
Boost clock
2,505 MHz

AI & compute performance

FP32
91.1 TFLOPS
RT
210.6 TFLOPS

Power & physical

TDP
300 W
Power connector
1x 16-pin
Form factor
Dual-slot
Cooling
Active
Length
267 mm

Interconnect

Interface
PCIe 4.0 x16
NVLink
No
Infinity Fabric
No

Virtualization

MIG
No
vGPU
Yes

Media engines

NVENC engines
3
NVDEC engines
3
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
INT8
Yes

RTX 6000 Ada frequently asked questions

Common questions about NVIDIA RTX 6000 Ada Generation pricing, specifications and workloads on HexGrid Cloud.

How much VRAM does the RTX 6000 Ada have?

The RTX 6000 Ada has 48GB of GDDR6 ECC memory with 960 GB/s of bandwidth.

Is the RTX 6000 Ada good for LLM inference?

Yes. Its 48GB memory capacity and fourth-generation Tensor Cores make it well suited to many LLM inference workloads.

Does the RTX 6000 Ada support FP8?

Yes. Its fourth-generation Tensor Cores support FP8 AI computation.

Does the RTX 6000 Ada support NVLink?

No. The RTX 6000 Ada does not support NVLink.

How much does an RTX 6000 Ada cost per hour?

Hourly pricing varies between GPU cloud providers. HexGrid compares available RTX 6000 Ada offers.

Other GPUs on HexGrid Cloud

If the RTX 6000 Ada isn't the right fit, these are also available to deploy.

Browse every GPU we rent

Start training on the RTX 6000 Ada today

$0.88 per GPU hour, billed per minute, running in about a minute.

Specification sources