Computers · datacenter GPU
Nvidia A100
Hacker Approved
Long-lived datacenter GPU that people still run for AI and compute work years after release; the most common caveat is that it can be unreliable under long runs.
Summarized from 14 comments.
Long-lived datacenter GPU that people still run for AI and compute work years after release; the most common caveat is that it can be unreliable under long runs.
- recommend
- 10
- criticize
- 4
- months recommended
- 8 of 13
Why people love it
Reliable after six years of use · 2
A100s are 6 years old at this point and still in heavy use.
Oct 2026 · commentA100s are 6 years old and still a reliable work-horse
May 2026 · commentViable cheaper alternative to Mac Studio · 2
that's not more expensive and certainly can't be slower than a Mac Studio
Aug 2026 · commentMaybe two/four of the older A100/A6000 could do the trick too.
Nov 2025 · comment
What people warn about
Unreliable over long runs · 2
it was hard to keep it running for more than a few days
Jan 2026 · commentmy numbers have GPUs faring even worse
Jan 2026 · commentExpensive to buy or rent · 2
Nvidia A100 (80GB VRAM) apparently cost $15,000 or so
Jun 2026 · commentWhat we're paying for five old 80GB A100s is criminal
Apr 2026 · commentResale value collapsing toward scrap · 2
there's a massive A100 liquidation sale that happened recently
Oct 2026 · commentThose A100s aren't far from being sold for scrap.
Jun 2026 · comment
Long-term owners
A100s are 6 years old at this point and still in heavy use.
Oct 2026 · comment
A100s are 6 years old and still a reliable work-horse
May 2026 · comment
What hackers pick instead
- NVIDIA H100People recommend it for top AI training and cheap rental access, though critics cite high purchase cost, power draw and failure rates.20recommend
- NVIDIA Tesla V100People recommend the used V100 as cheap VRAM and FP64 compute for local AI and physics work. Caveats: no bfloat16, loud fans, high idle power.16recommend
- AMD Instinct MI355XStrong memory and performance per dollar for large AI inference and training; critics say its software stack lags and it trails Nvidia's B300 on some benchmarks.14recommend
- Intel D270 GPU AI AcceleratorRecommended for holding many resident AI models in 480 GB of LPDDR5X memory. Critics say its low memory bandwidth makes it slow, and it is very expensive for individuals.10recommend
- NVIDIA B200Recommended for top raw AI throughput, a maturing software stack and low cost per token. Main caveat: rented B200 instances are scarce or pricey.9recommend
Every quote
The 13 most detailed of 14 comments. Each links to the original.
that's not more expensive and certainly can't be slower than a Mac Studio
Recommends · Aug 2026 · in “Qwen3.8-Max: A New Bar for Coding and Cowork”
Maybe two/four of the older A100/A6000 could do the trick too.
Recommends · Nov 2025 · in “Asus Ascent GX10”
there's a massive A100 liquidation sale that happened recently
Mixed · Oct 2026 · in “US Reps Lead Democratic Resolution Opposing A.I. Bubble Bailout for Big Tech”
A100s are 6 years old at this point and still in heavy use.
Recommends · Oct 2026 · in “Fiber lasts decades, GPUs last three years: the AI boom isn't the dot-com boom [video]”
A100s are 6 years old and still a reliable work-horse
Recommends · May 2026 · in “The AI bubble isn't like the internet bubble”
Nvidia A100 (80GB VRAM) apparently cost $15,000 or so
Mixed · Jun 2026
Those A100s aren't far from being sold for scrap.
Mixed · Jun 2026 · in “DeepSeek open-sources inference optimizations with 60–85% faster generation [pdf]”
it was hard to keep it running for more than a few days
Criticizes · Jan 2026 · in “Keeping 20k GPUs Healthy”
What we're paying for five old 80GB A100s is criminal
Criticizes · Apr 2026 · in “Apple approves driver that lets Nvidia eGPUs work with Arm Macs”
my numbers have GPUs faring even worse
Criticizes · Jan 2026 · in “Keeping 20k GPUs Healthy”
A100s are going to continue making money per hour until 2030.
Recommends · Jul 2026 · in “Nvidia, CoreWeave, and Nebius: Inside the Circular Financing of the GPU Boom”
Looks very playable to me.
Mixed · Dec 2025 · in “IBM CEO says there is 'no way' spending on AI data centers will pay off”
they'll still be slow and a little dumber than what you're used to
Criticizes · Jun 2026 · in “Ask HN: MacBook vs. Dedicated GPU for LLM”
Counts from 1,795,640 Hacker News comments, Nov 2025 to Nov 2026. How we count · Report a mistake