Tootfinder

i’m reviewing a paper on reducing energy costs in large model training and it keeps slinging words like optimize and optimization around and calling other approaches suboptimal and i feel like i would be kind of an old crank if i were to ask if optimality is on the table here (it is not)
EDIT: hold on, maybe it is

'This is the dark side and true meaning of "business optimization." The optimal business pays its suppliers and workers nothing, and charges its customers everything it can. Obviously, businesses need to settle for suboptimal outcomes, because workers won't show up if they don't get paid, and customers won't buy things that cost everything they have⹋.
⹋ Unless, of course, you are an academic publisher, in which case this is just how you do business.'
…

»Rückgrat — Python lehnt Millionenförderung durch Trump-Regierung wegen Anti-Diversitäts-Auflagen ab:
Die Python Software Foundation hätte mit 1,5 Millionen US-Dollar die größte Förderung in der eigenen Geschichte erhalten sollen«
Gut dass Python sich klar gegen Trumps Einkaufen von Propaganda wehrt auch wenn es Finanziell suboptimal ist. Nun können andere Sponsoren zeigen wie Weltoffen die wirklich sind.
:python:

Python lehnt Millionenförderung durch Trump-Regierung wegen Anti-Diversitäts-Auflagen ab
Die Python Software Foundation hätte mit 1,5 Millionen US-Dollar die größte Förderung in der eigenen Geschichte erhalten sollen

There's still room to tune - shallow memory throughput is definitely suboptimal due to latency issues I need to chase - but with deep memory ngscopeclient ThunderScope is getting some pretty impressive performance.
2 channels @ 50M point memory depth (100M points per trigger) streaming at 7.5 WFM/s over LAN from across the building through a router. 40Gbase-SR4 from client to core switch and from core switch to router, then back to core switch, then 10Gbase-SR to the machine host…