RHEL9 Likely To Drop Older x86_64 CPUs, Fedora Can Better Prepare With "Enterprise Linux Next"

[jabl]

VMWare has already sunsetted anything prior to 2010 (Westmere) due to missing instructions for virtualization. I wonder how much virtio stuff for KVM can be taken out for anything less than IvyBridge or Opteron Bulldozer.

KVM has always required hardware virtualization support. Now that such hardware support is ubiquitous it certainly makes sense for VMWare to follow suit and drop support for legacy hosts.

[Sonadow]

That kind of stinks.

I have a dual Xeon workstation that belongs to the Ivy Bridge era. It still runs rings around today's consumer hardware by virtue of its 192GB memory and dual processor setup, and is only beaten by the Core X and Threadripper HEDT families.

In addition, I have three low-cost Apollo Lake and Gemini Lake laptops that I bought from China at extremely low prices, and I fully intend to get a few more as spares due to their prices. They are excellent for daily computing such as email, writing, light Wiresharking / TCPdump and Chrome Remote Desktop (only for controlling another person's computer, not the other way round since it does not work on Wayland). And both Lakes do not have AVX. Those laptops are currently running Debian 10 on Wayland/

If Fedora is going to mandate AVX or AVX2 it's going to cut out a whole bunch of hardware that is fully capable of running a modern Linux distribution without performance issues. Especially on my dual Xeon.

[carewolf]

I think the cutoff should be AES support -- one generation before AVX (using Intel as the metric) which means anything from around 2010+ should be good enough.

That's the same as a SSE4.2 cutoff. And would work on all currently produced mainstream processors.

[pal666]

this circus of chasing slighly less ancient hardware is such a waste of resources. one would think redhat has engineers who are able to defer binding of codegen options to host cpu to install time or run time

[pal666]

My tough exactly. Impressive how chill the comments are overhaul. If it was Canonical, people had already ripped their panties and called Shuttleworth the antichrist.

because noone of commentators is using rhel on their junk intel laptop, so it doesn't matter to anyone?

[pal666]

Enabling the AVX2 compilation flag for the code which was not written to take advantage of this extension won't automatically increase performance. Just saying.

i'm glad you are not writing compilers

[sfx2000]

Just drop for CPUs without SSE4.x.

Profit!

No doubt - much like Apple dropping support for GPU's that don't support their Metal API - so Sandy Bridge and earlier are no longer supported by MacOS.

Spouse's MacBook Air 2011 got caught in that trap - at least with RedHat it's still useful... I suppose after 9 years of use, buying a new MacBook Air to replace it is reasonable to some, esp Apple

[sfx2000]

Came here expecting people to be defending their ancient hardware (like usual with these articles) and wasn't let down.

As others have said, RHEL is for servers anyway, not outdated and/or low end consumer machines.

Not really...

There's a lot of boxes still in the data center that run AMD64 w/o AVX - and these are more recent 22nm/14nm Intel SoC's - Rangely is a good example here, as a network oriented silvermont core. See these as cold-storage and edge computing platforms mostly - and being from a RedHat shop, one does like to keep things in a common environment.

Silvermont/Airmont - They are comparable to Westmere big cores, instruction sets...

I suspect there are a lot AMD's in a similar boat...

[SofS]

There seems to be some sort of misconception here. There is no way they do not know what families have or not some specific instruction set. On the contrary, for a serious organization like RedHat now part of IBM to consider it then it must mean that because they know their clients they know what instructions to target in principle.

The catch being what Michael pointed out, the QA related issues. And this is where this new build feedback tool makes sense, ideal for a QA test bed. For that matter I could easily see Fedora repositories mirroring the current ones except for enabling optimizations to the target instruction set they want to test, in a dumb manner, it will be for testing anyway, just replace one repository for another and give it a go to see what happens. First they will be able to look at the build problems related to enabling all these optimizations, then they could look for test user feedback. After a while, with the actual feedback on hands, then make a more qualified decision.

[StanGenchev]

I agree that the minimum requirement should be rised but the baseline should be an instruction set/feature of all CPU's from year 201x and upwards and not AVX/AVX2 which is found only in mid to high-end CPU's.

Both Intel and AMD are selling brand new processors which lack AVX support. Are they seriously trying to tell me that my brand new laptop with a last-gen Pentium CPU is "too-old" to run Fedora/RHEL/CentOS?

The company I work at, the companies at which many of my friends work at and in my personal life, RHEL is being used for both Workstation and server, so the statement that RHEL is server-only is quite ridiculous. And for anyone who might say "Just because YOU use it for that, does not mean it is MEANT to be used for that!", I would ask you to go to RedHat's website and see for yourself that RHEL is marketed as both Workstation and server.

So instead of fighting and splitting into "support everything" and "support only the newest high-end" groups, let us find a good middle ground. Personally I think that SSE4.2 is a good minimum requirement as there are no CPU's made after around 2012 that do not support it (at least not that I'm aware of, correct me if I'm wrong).