You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I note a closed issue regarding Arm support - #310 - so I completely understand if this issue gets closed for the same reason, but I'm asking essentially the same question from a different direction.
I have a small number of NVIDIA Jetson devices lying idle at present - including the AGX Orin Developer Kit, which has 32GB of shared RAM, a 12-core Arm Cortex-A78AE processor, and an Ampere GPU with 2,048 CUDA cores and 64 dedicated Tensor cores. It's a bit of a beast, in other words - delivers a claimed 275 TOPS for sparse networks.
I'd love to get the thing on the AI Horde as a worker, along with a couple of older Jetson devices, but I can't if there's no Arm support.
I'm unlikely to have time to dig into this myself any time soon, so I thought I'd throw the suggestion into the ring - but, again, I completely understand it's probably a pretty hefty undertaking and would be a very low priority feature request even if someone had the inclination to take it on.
The text was updated successfully, but these errors were encountered:
I note a closed issue regarding Arm support - #310 - so I completely understand if this issue gets closed for the same reason, but I'm asking essentially the same question from a different direction.
I have a small number of NVIDIA Jetson devices lying idle at present - including the AGX Orin Developer Kit, which has 32GB of shared RAM, a 12-core Arm Cortex-A78AE processor, and an Ampere GPU with 2,048 CUDA cores and 64 dedicated Tensor cores. It's a bit of a beast, in other words - delivers a claimed 275 TOPS for sparse networks.
I'd love to get the thing on the AI Horde as a worker, along with a couple of older Jetson devices, but I can't if there's no Arm support.
I'm unlikely to have time to dig into this myself any time soon, so I thought I'd throw the suggestion into the ring - but, again, I completely understand it's probably a pretty hefty undertaking and would be a very low priority feature request even if someone had the inclination to take it on.
The text was updated successfully, but these errors were encountered: