Printable View
32m pi and everest don't count, nor does anything cooler than 60C at the cpu ;)
Test as avg joe with 16g no fan on ram and something like an H70 with 1 fan in prime 95 blend 3 hours with 15gig in use inside a case closed up and show me the error of my ways :D
Don't forget to crank that cpu up to, we want that IMC nice and helter skelter on the edge.
Your dommies are not 4gb dimms:p:Quote:
Originally Posted by bassplayer
I have two sets of the 2133 8-9-8 in the box and they are stable (prime 95 blend) at 1866 8-8-8-24 at 1.65v, but those sticks are well out of the price range of the vengeance kit.
The vengeance kit is just a cheap kit and I don't think you should really expect to much out of them, I mean you can pick up 16gb for about $150, (or less if second hand) and that is great deal for those looking for quantity over quality.
Mmm 1600mhz Snipers indeed don't pass memtest at 1866 8-9-8 (me starts to eat one sock), only 4Gb dimms I have here that pass memtest at 1866 8-9-8 are the Geil Evo Corsas
Attachment 119863
Attachment 119864
Ofcourse we are comparing a rated 1600mhz dimm with a 2133Mhz one, whihc is kindda apples with oranges :) RipjawsX might have done the trick...
agreed in case versus on a test bench / mobo box is much different when testing stabilityQuote:
Originally Posted by chew*
Bumping this thread.... Very curious how Ket got with his Ripjaws, if Chew was spot on or Ket managed to suprise.
Fail at DDR3-1866 8-9-8-X ? Wow. That's really bad! Maybe good el cheapo ram for low end PC or HTPC or something but this stuff shouldn't have the connotations of performance or overclocking at all associated with it.
Are you talking about 4gb dimms Cryptik ?
This is how i run my 1866mhz vengeance kit (hynix). I use 1.605v on the ram.
Too bad this bios doesnt have half multis because then i could run the cpu-nb @ 2812 instead (12.5x225) because it wont do 2925 (13x225)
Anyways, i think there is a few more mhz or a little tighter (sub) timings left in theese sticks but im too lazy :p
strange...
my kit work well @ 9-10-9-27 2N 1.53v
passes ton of test....
Baahhh Corsair is in the need of thinking about those great mems they once had produced :( Comon Corsair!
My previous ram died on me (TR3X6G1600C7) but I got a Corsair Vengeance 12GB DDR3-1600 CL9 triple kit in return from Corsair.
I did read about this topic and my ram is also Nanya (8.16) so I thought my ram will propably suck big time.
But I was wrong, very wrong indeed.
Some results (quick testing memtest 86+ test 5, with increment steps of 50MHz on the ram):
999-24 1T 1.5V : 1700 max stable
999-24 1T 1.65V: 1755, 1 error
9-10-9-27 1T 1.5v: 1900 max stable (so the ram meets the also sold "1866MHz C9 Vengeance" ram)
9-10-9-27 1T 1.6v: 1950, 1 error
9-11-9-29 1T 1.6v: 2005 max stable
9-11-10-29 1T 1.65v: 2080 max stable
This is as high as I tested the ram. For my overclock I only needed 2000 on the ram.
With my 920 D0 it was allready quite stressful on the IMC, 1.47V was needed for QPI voltage.
Screen or it didn't happen :P >
http://i49.tinypic.com/2vwej38.png
Bumping this thread.
I just upgraded from my old deneb system to 3570k and Z77x-UD4H, so now i can really see what my BFRs can do.
Atm. im memtesting(4.20) them at 2400mhz 10.12.11.30 1T, rest of the subtimings on auto. 1.625 volt. 1.6v gave errors in test #4.
As ususal with my oc, im always looking for 24\7 clocks. So no crazy voltage and so on.
I might step down to 2200mhz since they can do that with 10.12.10.30 1T and 1.55v, possibly thighter too. I havent tested much around 2200mhz yet, just ran them at that settting. Tweaking 2400mhz first and see what i get.
Edit: They just completed 1 memtest pass @ 2400mhz :)
As discussed here in other threads and in extensive software bench testing at PC hardware review sites, there is no tangible gains with DDR3 RAM running above ~1600 MHz. on typical desktop PCs other than those with APUs where the GPU section can use the added bandwidth.
It's fun to play to see what your RAM will do but unfortunately there isn't any significant system gains even with overclocked CPUs other than in RAM benches which exaggerate the potential gains. If 1600 MHz. RAM were a system bottleneck, then increase RAM frequencies would show a significant gain as they did with DDR/DDR2 but not with the added bandwidth and frequency of DDR3.
Tighter timings make even less change than frequency increases with DDR3. When you increase the frequency, the real time for one cycle becomes less so when you lower the timing one cycle you've made a smaller change in real time than with slower RAM like DDR/DDR2. Unfortunately there just isn't anything of substance to be gained by using higher frequency than 1600 RAM on typical desktop PCs be they Intel or AMD powered, other than the entertainment factor of tweaking it to see what it will do.
People should buy what makes them happy but if you're buying faster than 1600 MHz. DDR3 RAM expecting to see a visible improvement in system performance, you will be surprised to find it simply does not happen. The RAM makers don't want you to know this but all you need to do is run your own test with real applications and see with your own eyes. Eventually we'll all be using APUs and then there should be a significant advantage to using faster RAM up to a point at least.
I know all that.
But even tho the difference is very small, it is a difference. And its not going to get worse. Its free performance, even tho it may just be 1%.
And it so happens that i DO play with IGP just now.
Ive also had this ram a couple of years. When i bought them BFR was very common on this and other ram and it was actually a good price on them. More value at that time would have ment generic 1333 ram, so why wouldnt i buy this ram instead?
Anyways, this is off topic. The topic here is Vengeance 1866 and what they can do.
Oh Noos :eek: AMDforMe is over here poopooing on faster RAM.IMO..faster RAM is better :P Tight timings are nice, too. :)
You are such a funny guy! Welcome to XS...where we run our ram at only stock speeds and undervolt our CPUs to save energy. :)Quote:
Originally Posted by schmuckley
Actually, there is a place here at XS for memory discussions of the nature that AMDforMe has contributed to...under General Computer Hardware there is a tab for General Memory. Here is the link:
http://www.xtremesystems.org/forums/...General-Memory
But Xtreme Bandwidth probably isn't the appropriate venue.
Ok ive tested abit more. I cant get 2400mhz stable. But im thinkin it might be something else than the ram itself. 11.13.12.35 2T, 1.65v is no different than 10.12.11.30 1T, 1.625v. Fails prime blend in seconds.
2200mhz 10.12.10.30 1T 1.55v is perfectly fine tho(again, i havent tried much else at 2200mhz yet.)
So it might be IMC or something.