This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2019-07-27
Channels
- # aleph (2)
- # announcements (7)
- # beginners (74)
- # clj-kondo (23)
- # cljdoc (3)
- # cljsrn (3)
- # clojure (42)
- # clojure-android (1)
- # clojure-uk (4)
- # clojuredesign-podcast (10)
- # clojurescript (4)
- # figwheel-main (19)
- # fulcro (19)
- # hoplon (4)
- # jobs (5)
- # juxt (8)
- # off-topic (5)
- # pathom (40)
- # perun (3)
- # shadow-cljs (56)
John Hennessy and David Patterson 2017 ACM Turing Award Lecture : Among many interesting points: - Mooreās Law running out of steam: for past five years, only 3.5% Y/Y increase in SPECint scores - but 65K speedup possible for Python programs (e.g., matrix mult) using techniques/technologies available today. - pointing at the need for domain specific hardware approaches (back to the 1960s-1970s again!) - integrated hw/sw design will lead to a new āgolden ageā for research https://www.youtube.com/watch?v=3LVeEjsn8Ts
Python folks already know that of course, and pretty much always call out to C libraries for the inner loops if they have heavy lifting to do. GPUs TensorCPUs and other deep learning inner loop acceleration hardware is widely used already, and at least a few people are looking for more such inner loops to turn into hardware.
any creative writing groups i could find for slack ?