• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

My Northwood so far at 2.75ghz EDIT: Back to 2.7 to be safe.

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.


<< Hey Tex, did you try running your memory at 3XFSB instead of 4XFSB to see if the memory was the limiting factor? I don't know which would be faster, 2.2@2.65GHz with RDRAM or 2.2@2.75GHz with DDR ram. We want some benchmarks!!! >>

Most high quality RDRAM it seems can usually hit 133fsb (PC1066) speeds, but some RDRAM can't. And really, you can't lower the fsb to RDRAM ratio to 3x unless the fsb is at 133, because if it's any lower, the RDRAM will be running as sub PC800 speeds. As for which would be faster? Good question. 😉
 
Nice overclock, I'm looking forward to building a Northwood system later this year myself. Keep us informed of any new info.
 
can't seem to find where i originally read it, but here's something from overclockers.com:



<< Something else rather interesting is that the maximum "official" voltage per VID is only 1.55V, only .05V more than the default voltage. Mobo manufacturers seem to be doing their own voltage mods to exceed that. The official Intel "don't you dare go above this" is only 1.75V, which is a lot lower compared to the default than we've seen in CPU designs before. >>



that's on the first page.

anyway, if you remember that the maximum safe voltage for 0.18 is 1.85v, 1.6 for 0.13 seems pretty damn high.

hey, you can overvolt as much as you want, but don't be surprised if it stops working after a month 🙂

Aelus
 
Damn, Tex. You just had to crap all over my new machine with yours, didn't you??? 😉

Now I feel so "inadequate" with my measly 2.4GHz 🙁

And yes, RDRAM doesn't seem to OC as well as other memory. At higher than 2.4 I got the dreaded Win2K/XP "systemroot/hive corrupted or missing" error at boot. I think I'll keep it clocked at 2.2 or 2.3 🙂

Well, at least I can fry an egg on my memory, and you can't! 😛 😉
 


<< Most high quality RDRAM it seems can usually hit 133fsb (PC1066) speeds, but some RDRAM can't. And really, you can't lower the fsb to RDRAM ratio to 3x unless the fsb is at 133, because if it's any lower, the RDRAM will be running as sub PC800 speeds. As for which would be faster? Good question. >>


You can set the ratio to 3X whenever you'd like, but lowering the ratio is only intented to be a test to see if the RDRAM is the limiting factor. If you can overclock more with ratio lowered then the ram is the limiting factor and you should wait for PC1066. 😉



<< could be rambus, but since the northwood is going to 533 mhz from 400mhz soon really doubt it.
 >>


Just because Intel come out with a chip that runs at 133MHz FSB (4 x 133MHz) doesn't mean your existing memory can run that fast. When Intel goes to 533MHz FSB CPUs then you'll have to get PC1066 if your PC800 memory can't handle the faster speed. 😉
 
wow tex, nice work.

Ya, we would all like to know what heatsink you are using.

was the chip retail?????

also, can you run sandra again and compare it to a high end (XP 1900+) Athlon?

thanks!

-Sam

WOW, what a pisser comp....intel is slowly fighting its way back into my respect
 


<< c'mon tex why do you think rambus was limiting factor here?

can tell you that the rambus was defintely the limiting factor on the i850

Both top scores were max voltage in bios. 1.65v in i850 and 1.77v on i845

sounds like limiting factor was mobo voltage settings as Both top scores were max voltage in bios

could be rambus, but since the northwood is going to 533 mhz from 400mhz soon really doubt it.

great to know you're so excited but facts please and nice overclock

peace and looking forward to benchmarks


if i am wrong and rdram not as overclockable please let me know as thats my next system
2.2 with abit th7ll raid >>



the reason I blamed the rambus was because if I lowered it to 3x I could go higher.
 


<< wow tex, nice work.

Ya, we would all like to know what heatsink you are using.

was the chip retail?????

also, can you run sandra again and compare it to a high end (XP 1900+) Athlon?

thanks!

-Sam

WOW, what a pisser comp....intel is slowly fighting its way back into my respect >>




Yes it was retail I just dont use the fan they give 🙂
 
Texmaster, since you've used both the I850 and 845 chipsets have you seen any noticeable difference due to DDR's lower latency? Do apps open any quicker and things like that.

In my Asus P4T-E, my 2.0A Northie does 2.2 at default voltage, 2.3 is about the max I can get stably. I'm leaving it at 2.2 now since it doesn't require any additional voltage.
 


<< Texmaster, since you've used both the I850 and 845 chipsets have you seen any noticeable difference due to DDR's lower latency? Do apps open any quicker and things like that.

In my Asus P4T-E, my 2.0A Northie does 2.2 at default voltage, 2.3 is about the max I can get stably. I'm leaving it at 2.2 now since it doesn't require any additional voltage. >>



Are you sure you are not running a Willimate P4? The voltages you describe were not available to me at that speed with my northwood but were with my willimate.

 
Nice overclock Tex.

Just a friendly warning, but according to aceshardware, running above 1.6V on a .13 process, is very risky if you want to keep your cpu longer than a couple of weeks.

Very true. I would be very careful overclocking your NW using anything above 1.7V. Here's what Aceshardware had to say:

...note that 1.7V is already to much for a 0.13-micron chip. In fact, did you know that these kinds of overclocks (voltage >110%, clockspeed >115%) are used to evaluate the durability of a CPU? CPUs that can last 10 years or more at their normal voltage, should be able to live a few weeks at these kinds of voltages. Don't use such high voltages (>1.65v) when overclocking because the finer the process, the fewer the number of atoms are used to make a transistor and the more likely it becomes that electromigration damages your chip.
 
Are you sure you are not running a Willimate P4? The voltages you describe were not available to me at that speed with my northwood but were with my willimate.


I think he meant 2.2ghz at default voltage(1.5v) not at 2.2v.
 


<< Yes I'm running a Northie, and at 2.2 gigahertz at 1.5 volts. >>



sorry about that Stan my bad.

I am going to drop back down after reading those same reviews to 2.7. There, I only need 1.65v to be stable.
 
Sweet overclock Tex.

It will be interesting to see how this new process will affect overclocking when voltage increases are required. It's pretty clear though that above 2.2Ghz the Northwood is peerless. AMD will have to wait until February until they can match the performance of Northwood. It's a shame SMT has essentially been disabled on the Northwoods. I wonder if buying now would be advisable to buying later...my thinking is that Intel will begin binning more accurately once faster processors are introduced.
 
I was actually hoping for a stronger overclock but I dont recommend ANYONE going up to 1.77v. Every time I did I got a hardware monitor error.

So in retrospect, the Northwood does do a better job of overclocking (at least for me) than Willy by about 200mhz.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top