New post: quantifying Gauss-Seidel's loop-carried dependency with OSACA, and a loop-unrolling fix that cuts it from 12 to 2 cycles/element. Three compiler experiments included to isolate the effect from vectorization.
👉 tinyurl.com/3h79t8fn
It's been a while ! I've just publish a new post: 𝐆𝐚𝐮𝐬𝐬-𝐒𝐞𝐢𝐝𝐞𝐥 𝐢𝐬 𝐟𝐚𝐬𝐭𝐞𝐫 𝐭𝐡𝐚𝐧 𝐉𝐚𝐜𝐨𝐛𝐢. 𝐎𝐫 𝐢𝐬 𝐢𝐭?
While keeping a high level p.o.v., I try to explain why Jacobi can be faster than Gauss-Seidel despite what the math says.
tinyurl.com/yme7dahk
There's still time to submit an abstract to FortranCon 2025. What are you waiting for?
events.fortrancon.org/event/…
Final submission deadline is September 12.
SC24 HPC creates! @thoefler is praising Fortran @fortranlang performance on a production code at the AI workshop for science 🤩!
Fortran is in blue, performance achieved not only on a CPU, but also on the GPU.
Long live to Fortran !
Full room for the 1st keynote talk by @thoefler on the AI for Science workshop #SC24.
We are at the age of computing, traditional HPC simulations on today’s supercomputers is not enough for the low resolution. Need AI models to push the boundary.
Fortran (@fortranlang) is now #8 at TIOBE Index and continues rising. With new compilers maturing quickly, Fortran will become a very nice language for numerical computing.
tiobe.com/tiobe-index/
🚨New post: "Creating a Conda package from Fortran using fpm and rattler-build" 🚨 In this post, I explore how modern tooling like fpm and rattler-build can dramatically streamline creating Conda packages from Fortran code.
samharrison.science/posts/co…@prefix_dev@fortranlang