RTX-3090-24GB vs RTX-PRO-6000-96GB
RTX-3090-24GB is the cheaper of the two: $0.09/hr at Salad, against $0.8/hr for RTX-PRO-6000-96GB at HyperAI — 8.9× the price.
Cheapest on each
Aligned facts
Same sheet, same order, both cards — including the gaps.
| Attribute | RTX-3090-24GB | RTX-PRO-6000-96GB |
|---|---|---|
| VRAM | 24 GB | 96 GB▲ |
| Bandwidth | ○ no answer | 1.79 TB/s |
| FP16 | ○ no answer | ○ no answer |
| FP8 | ○ no answer | ○ no answer |
| TDP | ○ no answer | 600 W |
| NVLink | ○ no answer | PCIe 5 (no NVLink) |
| Architecture | Ampere | Blackwell |
| Released | ○ no answer | 2025 |
A blank is not a zero. Where one card states a figure and the other states none, neither wins the row — the second number is unknown, and ranking an unknown as nothing is how a comparison starts to lie.
Every seller, both cards
One row per seller, cheapest tier, on-demand list prices only. Sellers of both cards lead, cheapest combined rate first.
| Seller | RTX-3090-24GB | RTX-PRO-6000-96GB |
|---|---|---|
| Lium | $0.252026-10-27JSON | $1.752026-10-27JSON |
| RunPod | $0.502026-10-27 | $2.492026-10-27 |
| Salad | $0.092026-10-26 | ○ no answer |
| Theta EdgeCloud | $0.142026-10-26 | ○ no answer |
| TensorDock | $0.202026-10-26 | ○ no answer |
| HyperAI | ○ no answer | $0.802026-10-26 |
| Shadeform | ○ no answer | $0.992026-10-26 |
| Beam | ○ no answer | $1.092026-10-27 |
| GPUhub | ○ no answer | $1.102026-10-26 |
| Vast.ai | ○ no answer | $1.402026-10-27 |
| Nova Cloud | ○ no answer | $1.562026-10-26 |
| Hyperstack | ○ no answer | $1.852026-10-26 |
| EcoHash | ○ no answer | $1.892026-10-26 |
| Enverge | ○ no answer | $1.952026-10-26 |
| Massed Compute | ○ no answer | $2.192026-10-26 |
| Spheron | ○ no answer | $2.352026-10-26 |
| Cerebrium | ○ no answer | $2.502026-10-27 |
| 3NV | ○ no answer | $2.502026-10-27 |
| Verda (DataCrunch) | ○ no answer | $2.572026-10-27 |
| Runcrate | ○ no answer | $2.592026-10-26 |
| Modal | ○ no answer | $3.032026-10-26 |
| Fal.ai | ○ no answer | $4.002026-10-22 |
What this page will not tell you
On-demand list rates only · spot and committed terms priced separately and not mixed in here · no performance-per-dollar figure appears on this page, because it needs a throughput benchmark we do not hold and dividing a price by a vendor’s own TFLOPS claim produces a number that looks authoritative and means nothing