CPUCores -- by the author of Dimmdrive



  • Well; here we go again. Wonder who's going to end up being suckered into this one.



  • Soo, it... sets the affinity for all processes to core #1 and lets them fight it off while keeping all other cores idle, despite the fact that many games are rather badly optimized for multi-core support so they don't need that many, the OS does a better job at scheduling processes and takes into account that the game is in the foreground, and that the CPU hasn't been a performance factor for the last 5 years or so?

    Well, it does... something, I guess.


  • sockdevs

    @Maciejasjmj said:

    Well, it does... something, I guess.

    yes. it separates idiots from their money


  • Discourse touched me in a no-no place

    Snake oil, again, then?


  • sockdevs

    worse than that, snakeoil was at least flamable.



  • @Maciejasjmj said:

    Soo, it... sets the affinity for all processes to core #1

    Presumably not the process for the game they're trying to "optimize", right? Othewise this makes double-no-sense.

    To be fair to the Steam community, the top few posts in the forum for this app are all asking, "where's the benchmarks?"


  • Discourse touched me in a no-no place

    Best part: On a modern CPU this will be counterproductive.

    Moving EVERYTHING to CPU0 will, on a crap infested shitbox (right now I have 30% across 4 cores running nothing but Javascript in Chrome), max out CPU0, which will heat up disproportionately to the other 4. On a crap-infested shitbox, the heatsink is probably below the thermal needs of the CPU - therefore CPU0 will go into throttling to keep temperatures down.

    The other cores will ALSO throttle down, because they share the same heat dissipation hardware.



  • @Weng said:

    Moving EVERYTHING to CPU0 will, on a crap infested shitbox (right now I have 30% across 4 cores running nothing but Javascript in Chrome),

    Ok your point is well-taken but...

    #HOW?

    My computer idles (a.k.a. "runs JavaScript in Chrome" at about 0.1%. How do you get a 30% idle state? What the fuck services do you have running?


  • sockdevs

    not everyone runs i7s

    my older (ivy-bridge i think) i3 laptop idles at ~30% usage, ~60% if i have a discourse tab open.

    no way i'm developing anything on it, but i'll do fine for watching youtube and browsing reddit.



  • @accalia said:

    snakeoil was at least flamable.

    If it runs everything on one core, maybe so is this?



  • There's no way even an i3 is idling at 30% unless you have tons of shit running on it.

    More likely, since it's a laptop, it's throttled WAY down to save power, and that's "30%, but the CPU's only running at like 20% speed" 30%.



  • Hell, my i5 is at 38% if both Chrome and WoW are running :)


  • sockdevs

    @blakeyrat said:

    There's no way even an i3 is idling at 30% unless you have tons of shit running on it.

    mine? chrome's about it.

    if i use IE it'll idle at 5% but then i'd be using IE. ick

    @blakeyrat said:

    More likely, since it's a laptop, it's throttled WAY down to save power, and that's "30%, but the CPU's only running at like 20% speed" 30%.
    yes, it does that. however when it throttles down it'll idle at about 70% of the lower CPU rate



  • I wonder if it would make Dwarf Fortress run faster with its multi-threaded game engine.



  • Have the ability to eliminate potentially inefficient and under-performing CPU "Hyperthreading" for a particular game

    WHICH WOULDN'T BE AN ISSUE IF YOU DIDN'T PEG EVERYTHING ON A SINGLE CORE. And AFAIK it isn't an issue anyway, as OS schedulers are generally aware of hyperthreads and don't schedule two demanding processes on the same physical core if it can be avoided. And also it brings properly-multithreaded games performance down, since a half-core is still better than no core.



  • @blakeyrat said:

    To be fair to the Steam community, the top few posts in the forum for this app are all asking, "where's the benchmarks?"

    And then the same community ends up being satisfied by a benchmark taken in Left for Dead 2 on ghastly lop-sided hardware no-one would ever use in practice. (Seriously; a crap dual core CPU paired with 16GB of RAM and a 960 Nvidia series card?)



  • Nah I read that thread, and a lot of people were calling it out as useless.



  • Well it is a good benchmarking procedure to have the other components of the system good enough not to interfere with the testing.

    It's still overshadowed by the fact that they didn't bother to pick one of the gajillion of games with standard performance tests, or launch a proper benchmarking tool. Also running the game in a friggin' window, which meddles with the scheduling.


  • sockdevs

    correted the @accalia in the title..... it was, ironically, annoying me.....


  • Discourse touched me in a no-no place

    I assume it's a combination of fuckbusted and old as dirt. I don't really do the home computer thing anymore.



  • Well I'm sure the 48 million people you're spamming on behalf of the Russian mafia appreciate your keeping the computer running anyway.



  • @Maciejasjmj said:

    It's still overshadowed by the fact that they didn't bother to pick one of the gajillion of games with standard performance tests, or launch a proper benchmarking tool. Also running the game in a friggin' window, which meddles with the scheduling.

    Eh, what? Why in the hell would windowed vs. fullscreen have any bearing on what the scheduler does?


  • sockdevs

    it's not an issue these days so much but fullscreening an app implicitly decreases the priority of all other processes on the system, leading to less jitter from processor context switching..

    or at least it used to do that on single monitor systems in the XP era. I imagine that is likely no longer the case in the windows 7/8 era



  • @tarunik said:

    Eh, what? Why in the hell would windowed vs. fullscreen have any bearing on what the scheduler does?

    It's leftover wisdom from like 10 years ago when that mattered.

    Dumb people have trouble getting stuff like that out of their heads. Maciejism probably runs defrag every week also.



  • @blakeyrat said:

    Maciejism probably runs defrag every week also.

    Hey! It makes the registry clean faster!



  • @blakeyrat said:

    It's leftover wisdom from like 10 years ago when that mattered.

    Yeah -- probably upwards of 10 years ago, even. Nowadays, video cards don't complain about doing 3D in any ol' viewport you ask 'em to, and even more than one at once.

    @blakeyrat said:

    Dumb people have trouble getting stuff like that out of their heads. Maciejism probably runs defrag every week also.

    Real filesystems don't need a defragger ;)


  • Discourse touched me in a no-no place

    @tarunik said:

    Real filesystems don't need a defragger

    Except when you're going to resize the filesystem to be smaller. Which is a rare operation indeed (especially if you don't install extra operating systems) and you're advised to actually do it by copying the data elsewhere, rebuilding the FS from scratch, and then copying the data back. Much safer that way.



  • @Maciejasjmj said:

    Well it is a good benchmarking procedure to have the other components of the system good enough not to interfere with the testing.

    In general, I agree that when running a scientifically valid test you should have as many constants as possible and only vary what you are actually measuring.

    What I'm opposed to is the fact that a system with lopsided specs is needed to demonstrate the software even has any positive effect at all. Basically, the benchmark was written against conditions that favor the software, instead of against the real-life conditions in which the software is expected to operate.



  • You still don't need a defragger because you can

    @dkf said:

    copy the data elsewhere, rebuild the FS from scratch, and then copy the data back



  • On NTFS, that does a much worse job than running one of MyDefrag's "monthly" scripts.



  • @ben_lubar said:

    You still don't need a defragger because you can

    You don't even need a computer because you could use a typewriter and a filing cabinet.



  • You don't even need a typewriter and a filing cabinet because you could just use two dried leaves and a dead bee. Creativity, people! Let's work smarter, not harder!


  • Discourse touched me in a no-no place

    @flabdablet said:

    Let's work smarter, not harder!

    Let's work dumber, not easier!



  • @flabdablet said:

    You don't even need a typewriter and a filing cabinet because you could just use two dried leaves and a dead bee. Creativity, people! Let's work smarter, not harder!

    I'm afraid my creativity falls short of imagining how two dried leaves and a dead bee could be used to achieve the desired effect.



  • You rub the bee juice on the leaves in the shape of words. But you have to write really small because the leaves aren't very big.


  • Discourse touched me in a no-no place

    And you don't get much juice from one bee.


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.