NG-SD-Plugins/PowerTools
NG-SD-Plugins/PowerTools
4
4
Fork 3
Code Issues 18 Pull requests 1 Projects Releases 24 Wiki Activity

Update 'FAQ'

NGnius 2023-10-30 20:32:09 +00:00
parent f93bb20145
commit fb2253a293
1 changed files with 116 additions and 122 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
FAQ.md

@ -32,22 +32,19 @@ Oh, you mean AMD PowerPlay! It's an AMD marketing term for their GPU (and APU) p
PowerPlay is a Table (hence the initialism PPT) of power limits.
Basically TDP.
## Why does PowerTools only let me select some values?
## Why does PowerTools only let me select some values? Why is X the default?
There's a few reasons for this, depending on the values in question, but it all stems from limits imposed by (or to protect) the hardware.
Most limits can be justified by the [specs of the Steam Deck](https://www.steamdeck.com/en/tech) or the allowed values for a specific kernel API.
## How do I reset to default settings?
Restart your Steam Deck. Nothing in PowerTools is persistent, unless you enable persistence (in which case: turn off the toggle, *then* restart your Deck). Or just press the Default button, but I'd imagine you would have found that already.
## Why is X the default?
I don't know, ask Valve. Most of the defaults are pretty logical, so none of them should be too objectionable.
## What does disabling SMT do?
The SMT toggle in PowerTools doesn't technically disable SMT.
Instead it disables every second CPU, since every group of two CPUs is one logical CPU core.
## Why does disabling SMT speed up some games?
I don't really know.
I don't really know. There is a bug before SteamOS 3.5 which caused cache issues when SMT was enabled, but there have been reports of some games still benefiting from SMT being disabled on SteamOS 3.5+ too.
My theory, which is backed by exactly zero research and experiments, is that since SMT increases performance of a single core by less than 100% (usually it's closer to 30-50%), that 150% performance gets split between two threads, effectively reducing each thread's performance to 75% when both threads are under heavy load.
It would then follow that disabling one of those threads would restore the remaining thread to 100% performance.
@ -109,10 +106,7 @@ Maybe. If it can run SteamOS 3, it can probably run Decky and PowerTools. Though
By the way, if you're the manufacturer of \<other handheld> and you want PowerTools to support it, send the device to NGnius.
Or, if not but you've got way more money than sense, buy an extra device to send to ~~a random stranger on the internet~~ NGnius.
In case you're seriously considering this half joke, [my contact info](https://github.com/NGnius/PowerTools/wiki/Contacts) is also on this wiki.
## Why am I stuck in an unhappy marriage?
Because your failure to communicate and/or get a divorce have left you trapped in a prison of your own creation.
In case you're seriously considering this half joke, [my contact info](https://git.ngni.us/NG-SD-Plugins/PowerTools/wiki/Contacts) is also on this wiki.
## This is really great, how can I support you?
I already make enough money, but if you want to help me buy new hardware to test on you can do so on [Liberapay](https://liberapay.com/NGnius/). Alternately, donate to some FOSS foundation and/or support a project that's mission critical, like the KDE Foundation or [core-js](https://github.com/zloirock/core-js#conclusion).