There are many ways to kill performance in a multithreaded program. My favorite one is false sharing, because you can get 10x performance just by moving some bytes around. So let's look into it. 00:00 - When multiple threads is not even faster 04:23 - CPU L1 cache primer 09:40 - Getting ALL the performance 15:35 - Looking at disassembler 20:29 - How to find cache line address 22:38 - false sharing in disassembly 25:00 - closing words

883

(3n+1)/2 · Oct 26, 2026 · 9:07 PM UTC

Math901 retweeted

(3n+1)/2

@Coolatz_Oct 26

Replying to @MathMath901

Now try this beauty! It's a KO!

205

(3n+1)/2 · Oct 26, 2026 · 6:50 PM UTC

Math901 retweeted

(3n+1)/2

@Coolatz_Oct 26

Replying to @MathMath901

Buenos días! Just came over this funny one 👇

1,483

Math901 · Oct 26, 2026 · 1:11 PM UTC

Math901 retweeted

Math901@MathMath901Oct 26

Good Morning earthlings.

996

ricardpe · Oct 26, 2026 · 6:05 AM UTC

Math901 retweeted

ricardpe@ricardpe1961aOct 26

Àlgebra. Problema 47

857

Logic Geometry, Information · Oct 25, 2026 · 5:34 PM UTC

Math901 retweeted

Logic Geometry, Information@LGcommaIOct 25

Hilbert swung so much to the other extreme as to not take consistency of the theory of with one constant-symbol 0 and one function-symbol S and only two axioms (S(x)≠0, S(x)=S(y)→x=y) for granted, in partc. rejecting the true-in-the-standard-model-argument! [Talk at ICM 1904]

Logic Geometry, Information@LGcommaIOct 24

In particular, there is the 'dialectical' aspect that Hilbert was in large part driven towards his formalist programme by the hostile reaction of the "community" to his novel use of 'ideal elements' in commutative algebra (in particular, quantification over infinite sets).

462