Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Quick Links
hpet urban legend!
drbaltazar
Member UncommonPosts: 7,856
ok a test called :wintimertester is going around claiming that hpet is better then invariant tsc etc!
its not!yes I know people will say but the number is higher!
trust me I tested!A LOT!
and my two cent?let intel processor handle things they are way better ms will ever be!
so disable hpet in both bios and os(the later via bcdedit)make sure netdma is enabled (for rss)
and make sure anything realted to offloading is disabled (no compatible with some other stuff .
don't be surprised here everything will look way more proper!my hunch?basicly I believe window cannot keep up with the processor .and with invariant tsc ?who would care
if you aren't using a cpu that do invariant tsc ?ignore this whole post!
Comments
*gets out popcorn*
I love these posts.
Indeed.
He did better with the first post and actually had some spacing. Downhill again with the 2nd post tho o.O
That is exactly the question I ask.
When I think of Urban Legends, there are a lot of various stories that pop into my mind. The nearest bathroom stall being the cleanest, Richard Gere and rodents, vampires in general, etc.
I certainly don't think of High Performance Event Timers.
Your best bet:
Stop playing with RGB color profiles, stop installing random driver "optimizations", stop using 3rd party non-official drivers for "performance testing".
Do use the latest 1st party drivers. Do keep them updated.
The odds of you beating an fully-updated system with default settings with various tweaks and oddball settings is low - not impossible, you can get some performance if your careful &/or lucky, but you probably won't eek out much over the default settings - a few single-digit percent if your lucky &/or careful.
The odds of you screwing something up and causing more harm than good; however, is pretty high. In fact, I would put it at several orders of magnitude higher at the odds of doing something bad over the odds of seeing any meaningful performance benefit.
Unless your doing something that requires you to be using a signal tracer and debugging your hardware/software at the electrical level on the wires (at which point a gaming site is really not your best outlet for information - you should be talking to IEEE), I'd say your pretty well wasting your time with this.
Should I go back and link all the other posts where you have made strange adjustments to your PC settings?
How many of those were still in place?
Maybe you legitimately had a problem with timers - it happens from time to time. But it's fairly uncommon. And given the frequency with which these kind of posts pop up where you have some off the wall setting that "helps PC Performance"... I'll let everyone else draw their own conclusions.
You think invariant time stamp counters are better than high performance event timers - you may very well be right, or maybe it's just a bug in your particular software or drivers, or maybe it's a result of some other setting being inadvertently off default.
Without a lot more information and backstory, this post is incoherent and useless at best, to misleading and potentially crippling at worst.
Citing a YouTube video with synchronization problems may not be the best indication of a very tight timing related problem. That could very well have been caused by network lag, Flash problems, browser issues, or any number of unrelated causes -- even the fact that it "cleared up" once you made some changes doesn't rule out any of those other possible causes. Doing a clean install after making some random BIOS changes also completely invalidates any hypothesis you want to test, because there are going to be hundreds of other factors that go into play with a clean installation, unless you are testing clean installation vs clean installation with just BIOS settings as the variable (which it doesn't sound like you started with).