This seems really specific, but if you're:
Uncategorized
1
Posts
1
Posters
4
Views
-
This seems really specific, but if you're:
✅ A security engineer
✅ UK based
✅ Looking for work
✅ Someone who loves catsThere's a remote role open now at Cats Protection that you might want to check out:
https://careers.cats.org.uk/job/799934 -
undefined oblomov@sociale.network shared this topic on
Gli ultimi otto messaggi ricevuti dalla Federazione
-
Giochi preziosi. Proteggere le Olimpiadi e reprimere gli scioperi
@anarchia
Non solo un’enorme speculazione e una nuova occasione di scempio ambientale. Le Olimpiadi attaccano anche il diritto di sciopero. All’inizio di dicembre infatti la Commissione di garanzia sugli scioperi ha pubblicato un invito rivolto alle parti sociali, cioè ad alcuni...Vedi l'articolo
-
me, swearing: OGG VORB IT
-
Current* conditions near Wasaga Beach, ON:
-
Carmine resistente: sabato “antisemitismo e antisionismo. usi e abusi” con amedeo rossi
@anarchia
Questo sabato nell’ambito del ciclo Carmine Resistente alle ore 18 allo spazio Aref, in piazza della Loggia 11 a Brescia, si terrà l’incontro con Amedeo Rossi, autore di “Antisionismo e antisemitismo. Usi e abusi”. Il libro di Amedeo Rossi appartiene...
-
Il mondo su misura (libro)
@libri - Il negazionismo scientifico può voler dire tante cose.
-
@alvan@social.lol Great, let me know when you come to me!
-
Writing an Optimizing Tensor Compiler from Scratch
Not everyone will write their own optimizing compiler from scratch, but those who do sometimes roll into it during the course of ever-growing project scope creep. People like [Michael Moroz], who wrote up a long and detailed article on the why and how. Specifically, a ‘small library’ involving a few matrix operations for a Unity-based project turned into a static optimizing tensor compiler, called TensorFrost, with a Python front-end and a shader-like syntax, all of which is available on GitHub.
The Python-based front-end implements low-level NumPy-like operations, with development still ongoing. As for why Yet Another Tensor Library had be developed, the reasons were that most of existing libraries are heavily focused on machine learning tasks and scale poorly otherwise, dynamic flow control is hard to implement, and the requirement of writing custom kernels in e.g. CUDA.
Above all [Michael] wanted to use a high-level language instead of pure shader code, and have something that can output graphical data in real-time. Taking the gamble, and leaning on LLVM for some parts, there is now a functional implementation, albeit with still a lot of work ahead.
