Originally Posted by Belial
I have to wonder about some of the sentiment being thrown around here.
Hardware has really come a long, long way, as well as software. I really wonder if software overclocking is as bad as it is made out to be anymore. Nowadays boards are basically sold on how good the included overclocking suite is, every single person at tomshardware buys their boards based on the software, and TH's reviews of motherboards ranks a board's ability to overclock not on it's VRM, but based on how far of a software overclock you can get (even if voltages are reported inaccurately 100% of the time on all motherboards no matter the brand and quality, and you can often use some motherboard oc programs on other boards).
I think it's a valid question to ask these days, if software overclocking really
is that bad. I mean people still say you should disable all powersaving options or that you should never go above 1.575 ram voltage on sandy bridge and ivy bridge, liquid pro being better than ultra, there's TONS of myths that are perpetuated because what used
to be true or based off some mythical intel spreadsheet that never existed. Remember the 'max voltage for ivy is 1.52v", haha.
I think you showed people like me that XTU might be better but nothing concrete. I haven't heard any reports, like someone saying "well i passed 24 hours of prime95 but failed in 5 hours of xtu' or any sort of comparison of the two, like 'well i know for sure i fail prime95 at .05v less than my current voltage but xtu passes just fine'.
...your point if what used to be true but isn't anymore is well taken...there was a time when all the memory you need 'is 640K
' ... that's why for commercial applications, we only use XTU (ditto for oc'ing and testing private systems)...in my experience, the XTU memory test results in a higher power draw and temps than prime95...typically 2 to 5 C higher, when I last ran a comparison (which is some time ago) though it depends on the system (ie speed and amount of RAM)
...I rarely use prime95 or the old, modified IBT these days anymore because I find XTU more challenging - and also what commercial clients can agree on
as a 'valid' test which we log on commercial builds before they leave...Then again, the XTU test result really doesn't say very much more than 'passed' or 'failed'...obviously, you want the 'green' passed
I ran some XTU tests last Friday before a weekend of record runs and there should be a log.
...a word of advice though...while I don't mind Asus AI Suite either if I'm in a hurry to test different scenarios, better don't run XTU and AI Suite at the same time.
..vCore can shoot up as those two can become additive on vCore...and I would know that because... Edited by Joa3d43 - 3/13/13 at 6:53pm