Running The AMD "ABBA" Ryzen 3000 Boost Fix Under Linux With 140 Tests

[phoronix]

Phoronix: Running The AMD "ABBA" Ryzen 3000 Boost Fix Under Linux With 140 Tests

Last week AMD's AGESA "ABBA" update began shipping with a fix to how the boost clock frequencies are handled in hopes of better achieving the rated boost frequencies for Ryzen 3000 series processors. I've been running some tests of an updated ASUS BIOS with this adjusted boost clock behavior to see how it performs under Linux with a Ryzen 9 3900X processor.

http://www.phoronix.com/vr.php?view=28265

[GrayShade]

The frequency graphs (e.g. Coremark) are confusingly labelled with "Megahertz, Less Is Better".

[Michael]

The frequency graphs (e.g. Coremark) are confusingly labelled with "Megahertz, Less Is Better".

Just ignore the "Less Is Better" label. All the other sensors (thermal, voltage, etc) all normally denote "Less Is Better" so that is the default for frequency sensor graphs. Just didn't have the time to update the code to handle otherwise before the article.

[bpetty]

Obe of the biggest improvements...

Found one spelling issue.

[Linuxxx]

Michael
Now that PTS can plot the maximum frequency, could You please also do another comparison with the newest BIOS of different CPU scaling governors?
I'm asking because I remember other Ryzen owners reporting that it was not possible to hit the boost frequencies with the performance governor.

BTW, thanks again for all the work You do for the greater good of Linux!
That's the reason why phoronix.com is the only website where I haven't blocked any advertisement since more than a decade now!

[oibaf]

What's the rationale for the order of the graph comparison? It looks properly order by absolute value, but the last 6 results are in a different order.
Also the geometric mean should really has the base value normalized to 100, it is 101 which it is not meaningful.

[Raka555]

Would love to see a 3900x vs 9900k @1080p gaming shootout

[AndyChow]

Wow, that's a lot of tests.

I still can't believe people complaining about not hitting boost speeds by 25 Mhz or 100Mhz. "I was cheated! My CPU is 0.5% less fast than it should be, whaaaa".

[caligula]

Wow, that's a lot of tests.

I still can't believe people complaining about not hitting boost speeds by 25 Mhz or 100Mhz. "I was cheated! My CPU is 0.5% less fast than it should be, whaaaa".

Well, for many this was the first time in years that they even considered AMD and immediately got burned. It was only like 50% cheaper and 200% more powerful. Pretty disappointing.

[rene]

1.7% or so are mostly in line with what ether youtubers found, too – the idle frequency spikes looked a bit suspicious though: https://www.youtube.com/watch?v=5Yz4cCncxVg