CPU designs must evolve to keep up with changing workloads.

YouTube video by Gauthier Dm - Science

With CUDA 13.2 that is now shipping, NVIDIA has provided official support for Red Hat Enterprise Linux compatible distributions/downstreams like AlmaLinux to CUDA. With this official NVIDIA CUDA support for these RHEL-compatible distributions, NVIDIA is also allowing the NVIDIA packages to be distributed directly from the OS package repositories...

About a year ago I posted about defer and that it would be available for everyone using gcc and/or clang soon. So it is probably time for an update. Two things have happened in the mean time: A tec…

YouTube video by GitButler

About jj is a Git-compatible version control system that is both simple and powerful. See the installation instructions to get started. Release highlights A new syntax for referring to hidden and...

NVIDIA's Olympus are the ARM64 cores found within the upcoming Vera CPU that will be paired with Rubin. Olympus cores are claimed to be twice as fast as NVIDIA's current CPU cores found in Grace and based on Neoverse-V2. Earlier this year the open-source compilers landed initial support for Olympus while now a proper CPU scheduling model has been upstreamed into LLVM 22...

Sometimes compilers can surprise and delight even a jaded old engineer like me

We learn that compilers have tricks to access memory efficiently

Why floating point maths doesn't vectorise like integers, and what to do about it

The art of not (directly) coming back: tail call optimisation

Inlining doesn't have to be all-or-nothing

A podcast about how the software industry got this way

Copy paste can sometimes be a good thing, at least if the compiler does it for you

Knowing how compilers call functions can help with design - and optimisation

Knowing when the compiler can't optimise is important too

When aliasing can prevent loop-invariant code motion

The compiler can move code outside of loops to speed things up

Pointer Arith (Using the GNU Compiler Collection (GCC))

Duplicating loops around can yield some decent optimisations

Hello, you fine Internet folks,

Learning when the compiler decides to unroll loops for performance

Compilers can rewrite loops to avoid expensive calculations

As GPU architectures rapidly evolve to meet the overcoming demands of exascale computing and machine learning, the performance implications of architectural innovations remain poorly understood acr…

Exploring the ways optimisers deal with loop constructs

How compilers avoid expensive division with multiplication tricks

Division doesn't have to be slow with some clever tricks

The ARM architecture has a cool feature, and compilers know how to use it