Nvidia raised the RTX PRO 6000 Blackwell MSRP to $16,000 for the 96 GB model, according to a Tom's Hardware report discussed on Hacker News. Pre-orders opened below $8,000 last year.
Model: RTX PRO 6000 Blackwell | VRAM: 96 GB | Price: $16,000 MSRP
What Changed in Pricing
The card's official price doubled from the initial pre-order range. The 96 GB configuration targets professional workloads that require large memory pools for model training and inference.
Hacker News tracked the story with 27 points and 13 comments. Users noted the jump occurred after initial low pre-order pricing.
How the Card Fits AI Workflows
The 96 GB capacity supports loading multiple large models simultaneously or running high-resolution diffusion and video generation tasks without heavy quantization. Professional RTX PRO branding includes ISV certifications and longer driver support cycles compared with consumer GeForce cards.
Benchmarks and Memory Advantage
No independent benchmarks appear in the source. The key published specification remains the 96 GB VRAM figure paired with the new $16,000 price. Prior-generation professional cards with similar memory typically carried lower MSRPs before the Blackwell refresh.
Pros and Cons
-
Pros
- 96 GB VRAM enables large-batch training and multi-model serving on a single card
- Professional driver stack and certification for CAD and simulation software
-
Cons
- MSRP doubled from early pre-order pricing
- Limited availability data at the new price point
Alternatives and Comparisons
| Feature | RTX PRO 6000 Blackwell | RTX 6000 Ada | AMD Instinct MI300X |
|---|---|---|---|
| VRAM | 96 GB | 48 GB | 192 GB |
| MSRP (current) | $16,000 | ~$7,000 | Server pricing |
| Target segment | Pro visualization + AI | Pro + AI | Datacenter AI |
The RTX 6000 Ada offers half the memory at roughly half the new Blackwell price. AMD's MI300X provides more memory but requires server platforms and different software stacks.
Who Should Consider This Card
Teams running local inference or fine-tuning of models above 30B parameters on a single workstation benefit from the 96 GB pool. Organizations already committed to Nvidia's professional ecosystem and needing certified drivers gain the most. Buyers seeking maximum tokens per dollar should compare against multi-GPU consumer builds or cloud instances first.
Bottom line: The doubled MSRP positions the card as a premium workstation option rather than a general-purpose AI accelerator for most independent developers.
Nvidia's pricing shift signals tighter margins on high-memory professional silicon as demand for large-context AI workloads grows.
Top comments (0)