2 ms·
Sounds fair. That's not the kind of machine I'd want as a development system though. And usually development systems are beefier than production systems. So cur
by amelius 10mo ago
Sounds fair. That's not the kind of machine I'd want as a development system though. And usually development systems are beefier than production systems. So curious how they'd solve that.
- Gracana 10mo agoYeah, it is quite specialized for inference. It's unlikely that you'd see this stuff outside of hardware specifically for that. Development systems for AI inference tend to be smaller by necessity. A DGX Spark, Station, a single B300 node... you'd work on something like that before deploying to a larger cluster. There's just nothing bigger than what you'd actually deploy to.
- transpute 10mo agoHBF, like expensive HBM, is targeted at AI data centers. The KAIST professor discussed an HBF unit having a capacity of 512 GB and a 1.638 TBps bandwidth. PCIe x8 GPU bandwidth is about 32GBbps, so HBF could be 50x PCIe bandwidth.