Ubuntu 24.04 LTS & Fedora 40 Continue To Trail Intel's Linux Performance Optimizations

[swastika]

IMHO one should believe every single of people who have "swastika" as their username. Doesn't matter if it is the left or right one....

My username is my real name (Uber had to apologize for discriminating based on names, maybe you should learn from them?) and it is considered an auspicious multi cultural symbol by millions of people for thousands of years (and continues to be the case today) before it was misused by bigots but hey if you want to continue that practice and discriminate people based on their name or religion and be a bigot, who cares what you believe.

[Just Some Noise]

I would like to see a comparison between CachyOS and Clear Linux. i suspect that Clear Linux still has the upper hand when it comes to performance, as its not only the -O3 flag that changes the performance. Will try Clear again. Usually i like to install my system myself, like Arch. But nearly 10% performance gain is really awesome.

Hopefuly somebody will nuke Jerusalem soon.

Please take your medication. People like you are a disgrace for germany and the world as a whole.

[arjan_intel]

The question I have is do these non-upstreamed optimisations have security implications? Given the multiple hardware level security vulnerabilities that have been publicised over the last half decade in the name of performance, and given that the little things add up in a Cloud environment... do these improvements increase attack surface or create new vulnerabilities which just haven't been identified yet because they're not implemented in the wider Linux ecosystem?

We have very very few of such non-upstreamed optimizations. And we do not compromise security for such patches.... that's just a bad idea
(and internal corporate rules on security also strictly don't allow that)

[woddy]

I did a test (just curiosity and has no statistical value), I installed the latest version of Clear Linux in dual boot with the distribution I usually use (Tumbleweed) on my Dell OptiPlex 7010 with integrated Intel graphics. I use office, web, some small video edits, music etc. classic desktop use and I haven't noticed any significant perceptible differences. Certainly for other uses it is more evident, but for general use I don't notice any differences.

[woddy]

Can you stop please?
We're talking about Linux here, not politics, have respect for others.
Thank you

[Rovano]

The question I have is do these non-upstreamed optimisations have security implications? Given the multiple hardware level security vulnerabilities that have been publicised over the last half decade in the name of performance, and given that the little things add up in a Cloud environment... do these improvements increase attack surface or create new vulnerabilities which just haven't been identified yet because they're not implemented in the wider Linux ecosystem?

What a performance. But what about safety in the whole process?

Certification?

It's just that some unstable distributions like Clear Linux or CachyOS don't stand a chance.
Just read user reviews.
And another thing is that many readers do not know how to interpret such tests correctly.

Apparently it was to make Clear Linux more visible.

You can do optimizations at all levels in your main linux distributions and you will see the performance.
But then who is to maintain it?

[bhptitotoss]

One one hand, Clear Linux is (Clear)ly the leader for low-level process-level optimizations. On the other hand, most infra use-cases are bottlenecked by networking to where it doesn't matter if you're running it on a few ARM instances or the fastest available Epyc monstrosity. Clear Linux clearly has a customer but I've not yet met them. I wonder if it's like special fx production or something?

[bgeneto]

Intel's ClearLinux boasts exceptionally fast boot times. One might wonder why other distributions haven't adopted similar optimization techniques. Fedora 40, for example, takes dozen of seconds while CL takes less then 3s to take me into a working Gnome desktop. (using Intel Optane)

[bgeneto]

I did a test (just curiosity and has no statistical value), I installed the latest version of Clear Linux in dual boot with the distribution I usually use (Tumbleweed) on my Dell OptiPlex 7010 with integrated Intel graphics. I use office, web, some small video edits, music etc. classic desktop use and I haven't noticed any significant perceptible differences. Certainly for other uses it is more evident, but for general use I don't notice any differences.

No difference? Not even in the boot time?!? 🤔

[Paradigm Shifter]

We have very very few of such non-upstreamed optimizations. And we do not compromise security for such patches.... that's just a bad idea
(and internal corporate rules on security also strictly don't allow that)

OK, thanks for the reply.