Remote GPU, inference & ML-systems jobs — the real numbers
Here is the honest picture most boards won't give you: of 142 open roles in this niche right now, only 12 are remote. GPU and accelerator work stays close to the hardware, the labs, and the data centers — so remote is the exception, not the default. These are the ones that exist today.
- open roles
- 12
- companies
- 5
- list salary
- 6 · $170K–$485K
- visa mention
- 2
- remote
- 12
Observed across current open postings, refreshed daily — not a survey. Salary band is drawn only from roles that publish a range. Salary breakdown →
The scarcity is structural, not a fluke of the moment. A large share of this field is kernel, silicon and cluster work that is tied to specific hardware, secure facilities, or co-location with a research team — none of which travels well over a VPN. When remote roles do appear here they cluster on the software-heavier end (inference, frameworks, tooling) rather than the hands-on-hardware end. If remote is a hard requirement, filter and sort below; if it's a preference, it's worth widening to hybrid, because the inventory is genuinely thin and refreshes daily.