alt Hacker NewsPied Piper vibes. As far as I can tell, this algorithm is hardly compatible with modern GPU architectures. My guess is that’s why the paper reports accuracy-vs-space, but conveniently avoids reporting inference wall-clock time. The baseline numbers also look seriously underreported. “several orders of magnitude” speedups for vector search? Really? anyone has actually reproduced these results?
Replies
NitpickLawyer • today at 10:35 AM
Apparently MLX confirmed it - https://x.com/prince_canuma/status/2036611007523512397
➕ show 1 reply
veunes • today at 9:49 AM
Classic academic move. If the authors show accuracy-vs-space charts but hide end-to-end latency, it usually means their code is slower in practice than vanilla fp16 without any compression. Polar coordinates are absolute poison for parallel GPU compute
➕ show 1 reply
Efficient execution on the GPU appears to have been one of the specific aims of the authors. Table 2 of their paper shows real world performance that would appear at a glance to be compatible with inference.