On GLM-5.2, Vera Rubin serves ~3x GB300's throughput per MW at ~165 tok/s per user, and sets the fastest GLM-5.2 result on our leaderboard at 408 tok/s per user
Vera Rubin serves GLM-5.2 at 408 tok/s per user, fastest on leaderboard
AISummary
Artificial Analysis reports that Vera Rubin serves GLM-5.2 at about 3x the throughput per MW of GB300 at roughly 165 tok/s per user. The same post sets Vera Rubin's 408 tok/s per user as the fastest GLM-5.2 result on the Artificial Analysis leaderboard.
Post on XView on X
Source: Artificial Analysis · x.comPublished