NVIDIA RTX PRO 6000
Hacker ApprovedRanked 3 of 42 graphics cards by recommenders
Recommended for local AI models and fine-tunes with 96GB VRAM and fast compute. Main caveat: very high price, and little gain over an RTX 5090 for inference.
Recommended for local AI models and fine-tunes with 96GB VRAM and fast compute. Main caveat: very high price, and little gain over an RTX 5090 for inference.
103 people gave an opinion
- recommend it
- 77
- criticize it
- 26
- approval share
- 75%
Recommended in 12 of 13 months
- November 2025: recommended
- December 2025: recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: recommended
- September 2026: recommended
- October 2026: recommended
- November 2026: not recommended (month in progress)
Why it carries the badge
- 77 people recommend itNeeds 10 or more, met.
- 3 recommenders for every criticNeeds 2 to 1 or better, met.
- Recommended in 12 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Runs large local models with 96GB
“That means that in the same number of slots you'll be able to put 3x as much VRAM and that's a massive advantage.”
Comment from 29 Oct 2026 in “GeForce vs. Workstation GPUs: Nvidia's Half Rate, Priced” “I have an RTX 6000 PRO with 96GB, and what I can run comfortably is Qwen 3.6 27B or MoE, Gemma 4 31B.”
Comment from 4 Jul 2026 in “Jamesob's guide to running SOTA LLMs locally” “The compute is insanely fast assuming you can fit what you need in 96GB of ECC GDDR7”
Comment from 21 Sep 2026 in “Samsung is expected to more than double output of its HBM4 and HBM4E DRAM” - Comment from 8 Apr 2026 in “GLM-5.1: Towards Long-Horizon Tasks”
- Comment from 11 Jan 2026 in “Don't fall into the anti-AI hype”
Fast training and inference compute
“The compute is insanely fast assuming you can fit what you need in 96GB of ECC GDDR7”
Comment from 21 Sep 2026 in “Samsung is expected to more than double output of its HBM4 and HBM4E DRAM” “pales in comparison to what a RTX 6000 Pro can do for compute/matmuls/training”
Comment from 25 Aug 2026 in “New Mac Studio with M5 Max and M5 Ultra” “can get ~250 tk/s decode with deepseek-v4-flash in native mixed precision”
Comment from 10 Aug 2026 in “Meta Muse Glimmer – open weights 30B local coding model” - Comment from 27 Jun 2026 in “Ask HN: MacBook vs. Dedicated GPU for LLM”
- Comment from 21 May 2026 in “Was my $48K GPU server worth it?”
Beats Mac Studio and gaming cards
“for the same money you’re far better off getting an RTX 6000, 5090, or R9700.”
Comment from 23 Jul 2026 in “Nobody knows what a used GPU cluster is worth” “the RTX 6000 will run circles around the Mac studio in just about every way”
Comment from 25 Aug 2026 in “New Mac Studio with M5 Max and M5 Ultra” Max-Q blower fits multi-card builds
“the Max-Q blower version of an RTX 6000 Pro Blackwell is easier to keep cool and also only needs 300 W at the cost of very little performance”
Comment from 22 May 2026 in “Was my $48K GPU server worth it?” “Max-Qs are the only card you can fit four of into this rig without PCIe extenders or cooking themselves”
Comment from 5 Jun 2026 in “The ways we contain Claude across products” “$50K will get you there very nicely, and on a 1600 watt PSU if you go for the MAX-Q versions.”
Comment from 23 Jun 2026 in “Unsloth GLM-5.2 – How to Run Locally” - Comment from 22 Mar 2026 in “Tinybox- offline AI device 120B parameters”
What people warn about
Each mark is one comment
Very high price, often thousands more
“if you just want stuff then you are absolutely better off not spending $20k on two of these cards”
Comment from 12 Jul 2026 in “I love LLMs, I hate hype” “the price for those GPUs is 5 figs even with the nvidia startup program”
Comment from 2 Sep 2026 in “My local model setup on an M4 Pro Mac Mini” “Looking forward to trying this out on my $10000+ workstation grade GPU that I need an equally expensive set up to run.”
Comment from 1 Jun 2026 in “Nvidia Cosmos 3” - Comment from 2 Sep 2026 in “A practical guide to running 8x RTX PRO 6000's”
Prices rising amid Nvidia dominance
“These GPUs are extremely inflated in price because Nvidia effectively has a monopoly”
Comment from 2 Sep 2026 in “A practical guide to running 8x RTX PRO 6000's” “they charge more than their competitors per flop and per byte because they’re the only manufacturer that can run CUDA”
Comment from 2 Sep 2026 in “A practical guide to running 8x RTX PRO 6000's” “The server edition has gone up $2K in the last couple of weeks alone”
Comment from 21 May 2026 in “Qwen3.7-Max: The Agent Frontier” - Comment from 21 Sep 2026 in “M5 Ultra Mac Studio Review”
Hard to pay off, may go obsolete
“I'd be better off paying 200 a month ... than buy up such a card and it be obsolete on purchase as newer better cards are always coming out.”
Comment from 4 Feb 2026 in “Qwen3-Coder-Next” “that card will _never_ pay for itself on a gpt-oss:120b workload vs just using openrouter”
Comment from 20 Dec 2025 in “Big GPUs don't need big PCs”
Other graphics cards people recommend
The four most recommended other graphics cards, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| NVIDIA GeForce RTX 3090 | 116 | 21 | 85% | 12 | 5 yrs14 owners | Approved |
| NVIDIA GeForce RTX 5090 | 102 | 37 | 73% | 12 | 1 yr3 owners | Approved |
| NVIDIA RTX PRO 6000This page | 77 | 26 | 75% | 12 | — | Approved |
| Nvidia GPUs | 67 | 67 | 50% | 11 | 6 yrs3 owners | — |
| AMD Radeon GPU | 48 | 29 | 62% | 11 | 2 yrs7 owners | — |
What people said
40 of the 140 comments that mention it, newest first. Each links to the original comment.
“That means that in the same number of slots you'll be able to put 3x as much VRAM and that's a massive advantage.”
“you could be paying off a set of 4 RTX 6000 cards that will run 4.1 Flash all you want”
“the NVIDIA hardware and the 6000 pro are priced less competitively”
“The compute is insanely fast assuming you can fit what you need in 96GB of ECC GDDR7”
“they charge more than their competitors per flop and per byte because they’re the only manufacturer that can run CUDA”
“These GPUs are extremely inflated in price because Nvidia effectively has a monopoly”
“Sure, let me just buy $60,000 worth of GPUs to run a *quantized non-frontier model*.”
“the price for those GPUs is 5 figs even with the nvidia startup program”
“RTX PRO 6000 96GB is really great for fine tunes (ai-toolkit) but doesn't outperform my RTX 5090 with inference by anything significant”
“pales in comparison to what a RTX 6000 Pro can do for compute/matmuls/training”
“the RTX 6000 will run circles around the Mac studio in just about every way”
“is not actually that expensive, and reports suggest that you can get 2k tok/s or so from a setup like this”
“can get ~250 tk/s decode with deepseek-v4-flash in native mixed precision”
“for the same money you’re far better off getting an RTX 6000, 5090, or R9700.”
“I almost bought four Sparks a couple months ago, but ultimately opted to buy two more RTX 6000 Pro Max-Q's.”
“if you just want stuff then you are absolutely better off not spending $20k on two of these cards”
“I have an RTX 6000 PRO with 96GB, and what I can run comfortably is Qwen 3.6 27B or MoE, Gemma 4 31B.”
“Not too long ago it was possible to put an 8x machine together for less than $100K USD”
“10x rtx6000 Pro in a large workstation is probably the way to go for someone wanting to run GLM5.2.”
“The Pro GPUs blast these out of water on performance but are expensive.”
“$50K will get you there very nicely, and on a 1600 watt PSU if you go for the MAX-Q versions.”
“Probably the most practical for cost/perf if you want on-prem as an individual.”
“Max-Qs are the only card you can fit four of into this rig without PCIe extenders or cooking themselves”
“Looking forward to trying this out on my $10000+ workstation grade GPU that I need an equally expensive set up to run.”
“the Max-Q blower version of an RTX 6000 Pro Blackwell is easier to keep cool and also only needs 300 W at the cost of very little performance”
“I've been seeing ~100 tps with 4-bit MiniMax 2.7 on two RTX 6000 boards”
“A pair of RTX6000 cards will give you a good performance boost due to tensor parallelism”
“one these gpus is a huge improvement over a laptop or a single 3090, you very quickly wish you had more”
“The server edition has gone up $2K in the last couple of weeks alone”
“96GB RTX PRO 6000 Blackwells would be the minimum to run a model of this size”
“they could had gone with the Max-Q version RTX PRO 6000 and only require 120V circuit”
“over a 5090 to an RTX 6000 Pro in my machine and be confident that my CUDA code will scale”
“I'd be better off paying 200 a month ... than buy up such a card and it be obsolete on purchase as newer better cards are always coming out.”
“gives you a machine that you can host locally and run very capable models”
“You're much better off getting an RTX 6000 or two, which is also gimped, but at least is much faster.”
“that card will _never_ pay for itself on a gpt-oss:120b workload vs just using openrouter”
“today you can buy a new, legit RTX 6000 Pro for about $7-8k, and get similar performance”
“the pricing for the rtx pro 6000 is surprisingly competitive with the gamer cards”
“This holds both for the sticker price, the supporting hardware price, the electricity cost to run it and cooling the room”
“put in a single rtx 6000 pro blackwell. That'd be a lot easier and probably a lot faster”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.