System76 Thelio Mira AI Linux Workstation
Shares rank 8 of 9 workstations by recommenders
People recommend it for running large AI models locally, with more GPU memory per dollar and faster token rates than a Mac Studio. Critics call it overpriced and raise PCIe bandwidth and thermal concerns.
People recommend it for running large AI models locally, with more GPU memory per dollar and faster token rates than a Mac Studio. Critics call it overpriced and raise PCIe bandwidth and thermal concerns.
20 people gave an opinion
- recommend it
- 6
- criticize it
- 14
- approval share
- 30%
Recommended in 1 of 13 months
- November 2025: not recommended
- December 2025: not recommended
- January 2026: not recommended
- February 2026: not recommended
- March 2026: not recommended
- April 2026: not recommended
- May 2026: not recommended
- June 2026: not recommended
- July 2026: not recommended
- August 2026: not recommended
- September 2026: recommended
- October 2026: not recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 6 people recommend itNeeds 10 or more, not met.
- 0.4 recommenders for every criticNeeds 2 to 1 or better, not met.
- Recommended in 1 of 13 monthsNeeds 4 or more, not 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
Faster token rates than Mac Studio
“This is a tier above the best Mac Studio and two tiers above the Spark.”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” “It's gonna have WAY faster token rates than that Mac Studio though”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” “It is more powerful than Apple.”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” More GPU memory per dollar
“More memory to store weight at a quarter of the price!”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” “I would rather take this option than a single H100”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory”
What people warn about
Each mark is one comment
Overpriced for its components
“The machine is overpriced, and I'd personally rather go for a second-hand workstation with server-grade components.”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” “That's a bitter pill to swallow. Most of this is the cost of the Nvidia card, and the RAM.”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” “50K without a proper CPU, mind.”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” - Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory”
Limited PCIe lanes and bandwidth
“the rather anaemic PCIE bandwidth is going to hurt”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” “There’s not enough PCIe lanes available to run everything without bifurcation.”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” “they're shipping RTX 6000s that are PCI lane-starved”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” Thermal problems with stacked cards
“isn't gonna work out thermally if you stack two of those workstation cards on top of each other”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” “would easily overheat if it's just two bog standard workstation cards on top of each other”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” Untested, unsatisfactory setup
“it's probably a fantasy setup, not built nor tested”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory” “To my mind, this isn't a very satisfactory setup.”
Comment from 11 Sep 2026 in “Thelio Mira AI Linux Workstation: 192 GB GPU Memory”
Other workstations people recommend
The four most recommended other workstations, 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 |
|---|---|---|---|---|---|---|
| Mac Pro | 49 | 92 | 35% | 11 | 16 yrs10 owners | — |
| Tinybox by tinygrad | 25 | 31 | 45% | 4 | — | — |
| Dell Precision | 23 | 12 | 66% | 9 | 10 yrs6 owners | — |
| Sun SPARCstation IPX | 12 | 5 | 71% | 4 | — | Approved |
| System76 Thelio Mira AI Linux WorkstationThis page | 6 | 14 | 30% | 1 | — | — |
What people said
19 of the 23 comments that mention it, newest first. Each links to the original comment.
“To my mind, this isn't a very satisfactory setup.”
“the rather anaemic PCIE bandwidth is going to hurt”
“it's probably a fantasy setup, not built nor tested”
“The workstation form factor only makes sense if you're locked out of cloud GPU allocation.”
“this is miss leading. actually it is 4gb ram a400 price.”
“would easily overheat if it's just two bog standard workstation cards on top of each other”
“50K without a proper CPU, mind.”
“they're shipping RTX 6000s that are PCI lane-starved”
“$42k and a toy CPU at that”
“The machine is overpriced, and I'd personally rather go for a second-hand workstation with server-grade components.”
“I would rather take this option than a single H100”
“This is a tier above the best Mac Studio and two tiers above the Spark.”
“There’s not enough PCIe lanes available to run everything without bifurcation.”
“That's a bitter pill to swallow. Most of this is the cost of the Nvidia card, and the RAM.”
“It's gonna have WAY faster token rates than that Mac Studio though”
“It is more powerful than Apple.”
“More memory to store weight at a quarter of the price!”
“isn't gonna work out thermally if you stack two of those workstation cards on top of each other”
“That's bonkers to buy this when newer flash models won't fit on this any more.”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.