Tiger Lake + Renoir On Ubuntu Linux For Battery vs. AC Performance

[phoronix]

Phoronix: Tiger Lake + Renoir On Ubuntu Linux For Battery vs. AC Performance

Given the recent Intel presentation alleging AMD Ryzen laptop performance being worse on battery relative to the AC vs. battery performance seen with Intel EVO notebooks featuring Tiger Lake processors, I ran a mini comparison on my side to see whether there is any merit to Intel's information when testing under Ubuntu Linux.

http://www.phoronix.com/vr.php?view=29722

[mlau]

So. intels Benchmark boost works as designed.

[fguerraz]

It's even worse with some models/manufacturers:

https://www.dell.com/community/XPS/XPS-9300-reducing-speed-when-battery-drops-below-15/m-p/7728138/highlight/true#M72416

[BNieuwenhuizen]

Did the Renoir performance change further with different battery charge as well?

[Michael]

Did the Renoir performance change further with different battery charge as well?

No apparent change at least for this particular laptop.

[mibo]

Others also checked Intels claims.
So, I learned that notebook manufacturers can configure the Ryzen boost mechanism to increase the frequency later to save energy...

https://www.extremetech.com/extreme/...ops-on-battery
German:
https://www.heise.de/news/Notebook-P...h-4969797.html

[andrei_me]

No apparent change at least for this particular laptop.

So Ryzen performance is the same/more predictable regardless of % of the battery, while in Intel the the performance degrades when below 50% of battery?

If so, the argument of Intel saying that AMD reduces performance to reach the same battery life applies more to them than AMD, as Ryzen keeps the same perf throughout the remaining battery

[Michael]

So Ryzen performance is the same/more predictable regardless of % of the battery, while in Intel the the performance degrades when below 50% of battery?

If so, the argument of Intel saying that AMD reduces performance to reach the same battery life applies more to them than AMD, as Ryzen keeps the same perf throughout the remaining battery

Right, at whatever battery percentages for Intel Tiger Lake appear to be different thresholds... The exact percentages are unclear due to presumably being within DPTF tables, etc.

[sykobee]

Rather than Intel's performance dropping when battery goes below a certain amount, I'd rather suggest that they are boosting performance - at the cost of battery usage - on well charged batteries.

This could be sensible - if the user is prone to only using the laptop off of battery power for short stints of time then this type of behaviour gives them full performance all the time, and it's sensible to use more power, at the cost of battery runtime.

This could be cynical - most reviews benchmark at full battery capacity, if they do a performance test on battery, of course.

Ideally this would be user configurable. Let the user choose Battery Performance Mode, or Battery Life Mode.

[blackshard]

Looking at the final geometic mean graph it is pretty clear that what Intel claims about AMD, is true for Intel itself

Since when on battery the trade off between performance and power consumption IS important, I still profoundly miss the power figures which may tell the whole story.