https://hackerpaper.kaipfister.de/p/1811.06992/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMThttps://news.ycombinator.com/item?id=19249680urn:hackerpapers:mention:1811.06992:19249680Mon, 25 Feb 2019 21:35:43 GMT<p>The authors trained ResNet-50 in 7.3 minutes at 75.3% accuracy. As a comparison, a Google TPUv3 pod with 1024 chips got to 75.2% accuracy with ResNet-50 in 1.8 minutes, and 76.2% accuracy in 2.2 minutes with an optimizer change and distributed batch normalization [1]. [1]: https://arxiv.org/abs/1811.06992</p><p><a href="https://hackerpaper.kaipfister.de/p/1811.06992/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.LG