NVIDIA B200
Shares rank 6 of 11 datacenter GPUs by recommenders
Recommended for top raw AI throughput, a maturing software stack and low cost per token. Main caveat: rented B200 instances are scarce or pricey.
Recommended for top raw AI throughput, a maturing software stack and low cost per token. Main caveat: rented B200 instances are scarce or pricey.
13 people gave an opinion
- recommend it
- 9
- criticize it
- 4
- approval share
- 69%
Recommended in 7 of 13 months
- November 2025: recommended
- December 2025: not recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: not recommended
- April 2026: not recommended
- May 2026: not recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: not recommended
- September 2026: recommended
- October 2026: recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 9 people recommend itNeeds 10 or more, not met.
- 2.3 recommenders for every criticNeeds 2 to 1 or better, met.
- Recommended in 7 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
Cheapest cost per generated token
“B200 is also the cost efficiency leader under updated run-cost estimates.”
Comment from 11 Feb 2026 on Hacker News “Nvidia B200s are the cheapest way, by far, to generate tokens”
Comment from 12 Feb 2026 in “GLM-5 was trained entirely on Huawei chips” Strong measured and raw performance
“Their measured performance on things people care about keep going up, and their software stack keeps getting better”
Comment from 15 Jan 2026 in “Furiosa: 3.5x efficiency over H100s” “While the B200 wins on raw FP8 throughput (~9000 vs 4614 TFLOPs)”
Comment from 27 Nov 2025 in “TPUs vs. GPUs and why Google is positioned to win AI race in the long term”
What people warn about
Each mark is one comment
Custom inference chips beat it on cost
“Even currently, Google's Tensor can do inference cheaper and more efficiently than a B200”
Comment from 14 Oct 2026 in “ASML's strong orders ease fears AI spending boom is running out of steam” “dedicated inference chips without an 80% profit margin rather than the B200”
Comment from 14 Oct 2026 in “ASML's strong orders ease fears AI spending boom is running out of steam” Rented instances scarce or pricey
“every time no such instances were available or they were more like 5x what you mention at the cheapest”
Comment from 9 Oct 2026 in “JetBrains reports revenue growth, net financial loss for 2025” “what's happening to those of us who need to rent 8xB200 to run a model like DSV4”
Comment from 23 May 2026 in “OpenAI Is Preparing to File for an IPO Soon”
Other datacenter GPUs people recommend
The four most recommended other datacenter GPUs, 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 H100 | 20 | 16 | 56% | 11 | — | — |
| NVIDIA Tesla V100 | 16 | 4 | 80% | 8 | — | Approved |
| AMD Instinct MI355X | 14 | 5 | 74% | 7 | — | Approved |
| Nvidia A100 | 10 | 4 | 71% | 8 | — | Approved |
| NVIDIA B200This page | 9 | 4 | 69% | 7 | — | — |
What people said
12 of the 13 comments that mention it, newest first. Each links to the original comment.
“Even currently, Google's Tensor can do inference cheaper and more efficiently than a B200”
“dedicated inference chips without an 80% profit margin rather than the B200”
“every time no such instances were available or they were more like 5x what you mention at the cheapest”
“a private H100SXM or B200 can go a long way”
“As long as they have the H100, B200, or B300, China won't be able to compete”
“what's happening to those of us who need to rent 8xB200 to run a model like DSV4”
“datacenter GPUs like B200 do fail regularly (within months in many cases)”
“Nvidia B200s are the cheapest way, by far, to generate tokens”
“B200 is also the cost efficiency leader under updated run-cost estimates.”
“Their measured performance on things people care about keep going up, and their software stack keeps getting better”
“you can't configure/tune vLLM on Spark, and then move it into a B200 and even expect it to work”
“While the B200 wins on raw FP8 throughput (~9000 vs 4614 TFLOPs)”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.