the high htt is giving the tccd an advantage.
Do some comparisons with 340-350x10 and 166div with bh-5(270-280) and max a64 tweaker settings. It stomps tccd :)
Printable View
the high htt is giving the tccd an advantage.
Do some comparisons with 340-350x10 and 166div with bh-5(270-280) and max a64 tweaker settings. It stomps tccd :)
And run the TCCD at 340-350 1:1? :confused:Quote:
Originally Posted by k|ngp|n
No way this TCCD can do that.
Neither can my watercooled CPU do 3400-3500MHz so i cannot test what you just said.
What makes you think BH-5 on a divider will stomp TCCD?
Did you try?
I'm not trying to argue that BH-5 at 270-280 is real fast but taking for granted it will annihilate all is a bit BH-5 biased, don't you think?
From the comparisations i've seen or done myself i can conclude that TCCD at 300MHz 2,5-3-3-7 1T can keep up with BH-5 at 270MHz 2-2-2-5 1T, perhaps you should give TCCD a little more credit?
:slobber: Never say never...............:Quote:
Originally Posted by Zeus
TCCD 1:1 -----------> 345,9 MHz !!!
:explode:
LOL, i was just talking about giving TCCD some more credit. :DQuote:
Originally Posted by HARDCORECLOCKER
Your right. 300x10 vs 270x10 is comparable, but the tccd still is using the higher htt on the mobo which gives it an advantage. In other words, 270mem speed max timings and 300mem speed max timings puts bh5 at a slight disadvantage. Run 340htt with 166 div and ram at 270 or so, and the bh5 scores benefit from running the higher htt, even on the 166div.Quote:
Originally Posted by Zeus
I'm purely talking from standpoint of 270 vs 300 forget the cpu divider and 1:1 to 1:1 comparison for now.
and you dont have to have 3400mhz cpu speed to run at 340htt with 166div and 270mem speed on bh5 ;). Keep in mind this wasnt with any ram timing tweaks, just 6-2-2-1 in bios. If there were, the score would be better no doubt about it as these and most bh5 can run much better using a64 tweaker.
Is there any ram period that can match these results?? Not 100% sure, but i dont think so. Tccd at 340 1:1 would have to have horrible timings im thinking and not come close. I am also not 100% sure if using 8 divider does anything to ram, but the fact that they still needed 3.5v to run that pifast tells me they were running full tilt.
fx55, neo2, hyperx3200 bh5
http://pic8.picturetrail.com/VOL256/...6/82964880.jpg
Maybe for the average user tccd is a viable option, but for top scores in benches and 3d mainly, bh5 still ownz IMO.
I still enjoyed your post very much tho :) I love comparisons like this and I know how much testing goes into it. thx.
Hehe, thanks man. ;)
I understand your argument about the TCCD benefit from the higher HTT but don't you think using a divider kills the advantage of the higher HTT?
To be honest i haven't tried any 3d benches so i cannot speak for that but i saw OPB doing a compare with BH-5 @ 11x273 vs TCCD @ 10x300 and Lobby high was about equal.
About your PiFast score, i definately wouldn't bet my virginity that TCCD @ 9x311 (=2800MHz) 1T 1:1 won't match or even top that score. :D
Like i said TCCD needs about 30 more MHz to match BH-5, so it will be a close call... :)
I'm glad you enjoyed the thread and thanks for your input.
I just thought of a scenario where tccd might have a chance :P
If you were to run say 360+htt with 166div and max timings on tccd or around 340htt with 183div on tccd. that would put it around 300-310 with tight timings, no? Is it possible to run some benches like that? You'll prolly need around 2.9ghz with lower divider I think.
Unfortunately, no.Quote:
Originally Posted by k|ngp|n
I just tried to boot at 320 1/2 divider, a no-go. :(
Now i understand why i could do 2.9GHz easy with BH-5 at 11 multi and anything over 315 with a 9 multi are unstable, it's just the CPU not liking anything over 315. :mad:
It's not the board, i've booted up till 420 on this board with a ClawHammer.
Time to explore the max of the CPU i guess. :)
:D Good idea - it mostly depends also on the board - my NEO2 does not go much above 300 1:1 even with the G.SKILL LA so my hopes are on DFI NF4.Quote:
Originally Posted by Zeus
Ya see it also in my BH-5 tests, I used MUSHKIN & CORSAIR and both maxed out exactly at the same speed of 275 MHz with 2-2-2-5, not 274 or 276, both exactly at 275.
This indeed is a sign for board is maxing out, not RAM so it it always important to check all the other components before blaming the RAM.....
:toast:
You know it's funny because I tried out the different settings..Quote:
Originally Posted by zakelwe
With cas=2, twr, trrd, twtr, trtw all at 2, with 8x and 6x I ran pifast.. Then, I put cas 1.5, and put everything else as tight as possible.. pifast was marginally lower.. It wasn't really that big of a difference.. I'm talking, practically nothing really worth noting..
If ya don't mine, write it down your Memory voltage, for more precision, because it's good to see the frequence & timing but with-out the voltage I can't really make any difference and tell witch mémory is better than other. :confused: Thank you for understanding :toast:Quote:
Originally Posted by HARDCORECLOCKER
I already hit 340MHz 2.5-4-3-7 1T, imagine what this will do with 2.5-4-4-8.Quote:
Originally Posted by Zeus
DDR690-700MHz is possible with some tweaking.
BH5 is faster and more efficient then TCCD, everyone knows that.
The fact is that TCCD tramples BH5 with the very HTT close to DDR700.