My 2.6C overclocking results!

aznskickass

Member
May 3, 2002
54
0
0
Hi all.

Just thought you guys might be interested in my o/c results...

2.6C Week 21 Malay chip (L321A), on an Asus P4P800 Deluxe mobo. HSF is a Zalman CNPS7000Cu. Only using 2x 256MB Kingston PC2700 ValueRAM atm (3:2 divider, 2.5-3-2-7 timings), waiting for the 2x 512MB Hyper-X 3500 to come in.

I am currently running it at 3397MHz 1.6V, which seems stable for everything *BUT* Prime95. For Prime95 to run at 3.4GHz error free, I need around 1.75V, which isn't exactly safe for 24/7 usage IMO. But I ain't too bothered with Prime stability, as everything else runs fine for me, so that's all that matters. Stock Vcore is stable up to 3.3GHz, again, for everything except Prime, which requires 1.6V at such speeds.

The highest speed I can run 3DMark at 1.75V is 3490MHz. I seem to hit a bit of a wall here, even 1.8V doesn't seem to help much, with 3DMark completion at 3501MHz hit and miss...

With the 9800 Pro overclocked to 455core/375mem I can get a 3DMark of 20400 - 20500.

I imagine once I get the Hyper-X 3500 in I'll have a shot at 22000 3DMarks. I hope I can anyway.

I'll keep you guys posted on how I go.
 

bacillus

Lifer
Jan 6, 2001
14,517
0
71
nice o/clock but suggest that you don't exceed 1.7V vcore if you don't want that cpu to suddenly die on you!
 

stevejst

Banned
May 12, 2002
1,018
0
0
nice o/clock but suggest that you don't exceed 1.7V vcore if you don't want that cpu to suddenly die on you!

It is very unlikely that Pentium would die suddenly because of temporary use of higher voltage. Check this.
It is more likely it will be damaged over time, due to excessive voltage. This processor is a lot better equiped to handle emergency than old generation of CPUs or AMD CPUs.
I don't know anybody that had a Pentium IV died on them, unlike AMD.
It looks to me that aznskickass has an excellent and stable setup whose only "weak" point is the CPU, if you can call a weak the one that goes that far. Who can blame him to try to find out how much he can push it?

Just you would have much better results using improvement in cooling than jacking up voltage (that is the real difference between Pentium and Athlon), so once you reduce the temperature you can actually apply 2.0+ voltage on it!
 

batmang

Diamond Member
Jul 16, 2003
3,020
1
81
hell, i can run my cpu vcore at 1.8v and its pretty much the same temperature as my current 1.55vcore, about 45 c FULL LOAD.
 

sman789

Banned
May 6, 2003
1,038
0
0
Originally posted by: batmang
hell, i can run my cpu vcore at 1.8v and its pretty much the same temperature as my current 1.55vcore, about 45 c FULL LOAD.

:Q
 

stevejst

Banned
May 12, 2002
1,018
0
0
That is a thermal control of the CPU, at some point gate clocking kicks in and you cannot go higher but the performance suffers and Prime95 will fail.
 

stevejst

Banned
May 12, 2002
1,018
0
0
There have been many reports of excessive voltage killing P4 processors.......
I noted the experience of people I know. Or should I say -- there is much less chance to have Pentium die on you than AMD CPU. But if you are using excessive voltage for extended period of time everything is possible. The thermal diode might be defective, the motherboard might be defective, the power supply might be defective. The CPU itself might be defective.
Nevertheless the safeguards Pentium IV has are a lot better than any other processor currently in use.
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
Originally posted by: stevejst
There have been many reports of excessive voltage killing P4 processors.......
I noted the experience of people I know. Or should I say -- there is much less chance to have Pentium die on you than AMD CPU. But if you are using excessive voltage for extended period of time everything is possible. The thermal diode might be defective, the motherboard might be defective, the power supply might be defective. The CPU itself might be defective.
Nevertheless the safeguards Pentium IV has are a lot better than any other processor currently in use.


Speaking from experince with early northwoods when we really didn't know the boundaries of the vcore and though that as long as temps were fine we were fine...I can say that catastrophic or instance death is rare instead you first start to loose ability to hold stabile at oc speed with often several step downs and sometimes all the way back to default still results in all sorts of BSOD and fatal errors in apps....

Early northwoods I would stay below 1.75v...Newer p4bs and p4cs I would stay below 1.7v....


 

aznskickass

Member
May 3, 2002
54
0
0
I only raise the Vcore to 1.75V or higher when I'm benching, only lasts a few minutes a session, don't think it will do the chip much damage at all.

As I said earlier, my 24/7 speed is 3.4GHz 1.6V, I don't think this would do the chip any harm.

Cheers.
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
Originally posted by: aznskickass
I only raise the Vcore to 1.75V or higher when I'm benching, only lasts a few minutes a session, don't think it will do the chip much damage at all.

As I said earlier, my 24/7 speed is 3.4GHz 1.6V, I don't think this would do the chip any harm.

Cheers.

I agree with that and like you thinkin!!!

Good luck and nice chip....I would love to step up to HT with the c chips but will hold off until I see more of these hitting 3.6ghz + ....

 

aznskickass

Member
May 3, 2002
54
0
0
Originally posted by: Overkast
azn, how long have you been OC'ing your 2.6c @ 3397Mhz with stability?

Almost two weeks now. Hasen't skipped a beat. I 'burned it in' at 1.65V for a few days, but then reduced it to 1.6V and it seems pretty stable so far.

 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
1.6v will be a nice safe range...I have been at 1.58vactual at 3.24ghz since april and never missed a beat.
 

aznskickass

Member
May 3, 2002
54
0
0
Yeah, IMHO, it really isn't worth it to get that last 100MHz with an extra 0.15V or whatever. In fact, I might even consider running it at stock Vcore 3.3GHz during summer (I live in Australia, so it's winter here atm).

I really doubt I would notice the difference between 3.4GHz and 3.3GHz, it's a ~3% reduction in clockspeed, and considering most applications don't scale perfectly with CPU speed, I'll probably only lose ~2% in overall performance.

Unless you are god, no one will notice such a difference.
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
I can run stock vcore at 3.15ghz but I really don't get much of a temp increase going from 3.15ghz and 1.53v to 3.24ghz and 1.58v. I could push it higher like 3.3ghz with 1.65v but I can't go past 180fsb since the ram cannot do better then 450mhz ddr with the setting as low as possible....

 

aznskickass

Member
May 3, 2002
54
0
0
That is actually quite a good increase in clockspeed with only 0.05V increase, 90MHz. I need an extra 0.075V just to get an extra 97MHz! Oh well, I can't really complain, they did say D1s don't scale as well with voltage (I've seem some that don't scale AT ALL no matter the Vcore).
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
But remeber that 97mhz took like 7-8fsb to get it while with my 90mhz it only took 5fsb. It seems that often the chipset and the mobo require a bit more juice as well when raising the fsb.

i would like to get a 2.6c as good as yours with a dcddr setup and I would be happy. The cpu speed would only constitute a 5% gain in overall speed but the fsb would be a 45% gain and the memroy performance could be in the 15-20% range, not to mention HT and I multitask big time and could really gain from that....


How is that hyper x 3500 stuff??? affordable??? Does it have some headroom left you think,??? Obviuosly you are running a 5:4 mem divider, right??
 

aznskickass

Member
May 3, 2002
54
0
0
I'm getting the Hyper-X 3500 tommorrow, so I'll tell you how I go. It's pretty 'affordable' in that it's cheaper than Corsair/OCZ stuff, but it's not exactly budget RAM either!

 

Overkast

Senior member
Aug 1, 2003
337
0
0
Yes, please do keep us posted! I'm on the verge of purchasing some OCZ 3500, but I'm interested in alternative solutions if they're comparable in performance.

Also, 1 more quick question for ya: What is the advantage of burning-in your CPU at 1.65v instead of 1.6v? Is it just a matter of burning off an extra bit of silicon for cooler temps?

**Edit**

Here's the specs of a system I'm about to build in 1 month from now (of course, all components subject to change based on continuous feedback and research):

-Chieftec Server Chassis (Aluminum w/ side window)
-P4 3.0c / 800FSB
-SLK-900U HS
-92mm Vantec Tornado (I don't care about noise)
-ASUS P4P800 Deluxe
-1 GB OCZ PC3500 RAM
-120 GB Maxtor HD (8 MB cache)
-DVD ROM
-CD-RW
-Floppy Drive
-GeoForce 4 Ti 4400 / 128MB (already own it)
-Allied TRUEPOWER 450W PS