3 ms·
Last time I checked SNNs still don’t work as well as traditional NNs. Why are they working on hardware before getting software to work?
by p1esk 3y ago
Last time I checked SNNs still don’t work as well as traditional NNs. Why are they working on hardware before getting software to work?
- johnnyjeans 3y agoTraditional NNs didn't work very well either, until we were able to run them at massive scales. That's why they didn't see much adoption or use until the late 2000s/early 2010s. SNNs are much more difficult to run at scale, but that difficulty is solved with analogue computers[1][2]. We don't have details, but I'd bet this is just Intel's crack at a modern analogue processor, with some design concessions specifically for running SNNs. [1] https://www.mdpi.com/2079-9268/13/4/55 https://www.mdpi.com/2079-9268/13/4/55 [2] https://www.mdpi.com/1424-8220/23/14/6275 https://www.mdpi.com/1424-8220/23/14/6275
- p1esk 3y agoLoihi is a digital chip.
- orbital-decay 3y agoSNNs are behind, but not that far behind, especially after recent astrocyte-related advancements, and if implemented correctly in hardware they consume a fraction of power. SNNs don't get nearly the same amount of attention as ordinary networks largely because they lack efficient hardware to run on (just like ordinary NNs until the advent of GPGPU). It's not exactly new hardware though, it's just a cluster of 1152 existing Intel Loihi 2 neuromorphic chips.
- p1esk 3y agoWhich “recent astrocyte-related advancements”?