i dont know if any one mentioned this already but PAT, a canterwood feature, isn't even a technology. It's simply a term for using better silicon. So PAT doesn't really exist.
Printable View
i dont know if any one mentioned this already but PAT, a canterwood feature, isn't even a technology. It's simply a term for using better silicon. So PAT doesn't really exist.
erm I think many will question this! PAT seems to be some ram optimizations that probably aren't inherent in certain silicon other than Intel possibly disabling it in the 865PE series.Quote:
Originally posted by oplix
i dont know if any one mentioned this already but PAT, a canterwood feature, isn't even a technology. It's simply a term for using better silicon. So PAT doesn't really exist.
Quote:
Originally posted by JNav89GT
erm I think many will question this! PAT seems to be some ram optimizations that probably aren't inherent in certain silicon other than Intel possibly disabling it in the 865PE series.
I think they would likely select the better silicon for the PAT enabled 875, since this "shortcut" to decrease latency is a little more difficult for the chipset. However, looking at 865's OC capabilities, I dont think there's much difference between them in terms of silicon quality
Yes, they do speed bin chips, and use the best for 875Ps. However, they do alter the chipsets at that point, to shave off clock cycles in the memory controller itself. They are physically different. If what you are saying were true, the two chipsets would give the exact same performance at the same clockspeed, oplix. This is not the case.Quote:
Originally posted by oplix
i dont know if any one mentioned this already but PAT, a canterwood feature, isn't even a technology. It's simply a term for using better silicon. So PAT doesn't really exist.
Dang macci...that is crazy.
I saw you mention, though, that your 875P at 300 FSB whipped your 865PE at 340 in 3DMark?
That seems funky to me...any idea what's going on there?
Macci, I just saw your 320+ FSB, looks like you almost maxed out that board from what anandtechs review shows it only gets up to 333. thats insane.
did you volt mod that board?
from the review it looks like it only gets 1.6 for Core Volts.
Zroc, Yeah the Albatron 865PE @340FSB w/ its slow bios does perform worse than a IC7/P4C800 at 300FSB. But thats only because of there silly slow BIOS.
yes it does seem slow :( compared to 4.1 GHz and much lower FSB....Quote:
Originally posted by macci
Zroc, Yeah the Albatron 865PE @340FSB w/ its slow bios does perform worse than a IC7/P4C800 at 300FSB. But thats only because of there silly slow BIOS.
http://users.pandora.be/JCviggen/p4pi.JPG
Yeah..and the SuperPi w/ albatron is actually quite good..3DMark is even slower (talking about 108Car hi at 4.2GHz etc.) LOL
IC7 at 274FSB gave 34sec btw ;)
I think I've got 36sec aircooled w/ P4C800+2.4C too.
Awesome scores macci, congrats.
What's your memory voltage on the IC7 @ 220Mhz, 2-5-2-2 (looking at your screenshot)?
Hmmm... *thinks about the ASUS* *remembers his and JC's P4T533 accident*
I'm pretty pleased with my IC7-G :)
LOL @ major sluaghter,
Quote:
Originally posted by macci
IC7 at 274FSB gave 34sec
Well see you could stay at 5:4... I had to go to 3:2 at those speeds :(
At 275 5:4 on the IC7 I could max out the mem timings/options and in lobby it gave me the same results as 285 more relaxed timings or 295 3:2 :(
Or just do this.Quote:
Originally posted by Zroc
Well, you could do that...you'd need to multiply the result times 2.
It's just easier to go FSBx1.6 than FSB/1.25x2. Plus, a lot of motherboards list their ratios as 1.6 or 1.3.
Take 5:4 Just multiply your FSB by the number on the right being 4, and then divide that by the number on the left, 5.
So at say a FSB of 250 gives you using 5:4 is
250 X 4=1000
1000 divided by 5=200mhz
Man that is a pain.... I guess no one here is a math genius! To get true Ram speeds do this, very simple:Quote:
Originally posted by TechTones
Or just do this.
Take 5:4 Just multiply your FSB by the number on the right being 4, and then divide that by the number on the left, 5.
So at say a FSB of 250 gives you using 5:4 is
250 X 4=1000
1000 divided by 5=200mhz
5:4 = Just multiply FSB x .8
3:2 = Just multiply FSB x .666
a pain?
imo all the other ways that have been explained are a pain.....its really quite easy to go 260 *4/5 = 208
Ive been doing since using the old 3:4 dividers on the PE chipset....mega mega easy way of doing it :thumbup: :D
I'm using the 326 x 4 / 5 method with my P4P800 :D
http://www.solidhardware.com/macci/p...p_326mem_s.gif
Now lets see a P4C800 do the same :D
Seriously the P4C800 is a bit faster clock to clock but it cant go much above 300 5:4 mem. So at 320FSB P4C (3:2) and P4P (5:4) are equal. Gotta do some more testing thou.
macci would you mind doing a pcmark benchmark for me? im curious to see your scores
Maaci
are you running a Prommy with a P4P800? Any reset switch problems where the board doesn't wait for the evap reach -33 but boots up *right* away.
Opp said there are problems with this 865 board too like mots of them and the Prommy.
macci,
what has been your favorite so far of the of 875/865 boards?
P4C800 and P4P800 are pretty much equal boards. P4P is obviously a bit cheaper and it requires 300+fsb 5:4 mode in order to compete w/ P4C800 at 300fsb 3:2. So if your going to for extreme performance w/ P4P800 make sure you've got some good ram (DDR500+)
what ram do you have? ocz pc3700 gold & whats your vdimm?
Kingston HyperX PC3000
Kingston HyperX PC3000:eek:
WHo'd a though??? What chips are on them??
Hi Ted :)
I have a 2.6c and P4P800 coming tomorrow. What ram are you using for Intel 865's?