How high should I set multiplier on new Pentium 965EE?

I just found a great deal on a 965EE, which is the very top my mobo (intel D955XBK) can handle. My last chip didn't have an unlocked multiplier, and I am wondering how high I should set it. Most places say it's safe to jump a few notches even with fan cooling. I have a liquid cooling kit for my processor (because of prior overheating problems).

Any thoughts?

Also, what should I check as I go? I got Speedfan to check CPU temps, and Prime995 to test it.

11 answers Last reply
More about high multiplier pentium 965ee
  1. Great deal better mean you picked it up for $20 or less. Most/all C2Ds are faster then this chip.

    Stock is 14, I doubt you'll be able to get it much above 16. It's already running at 3.73GHz, most of these chips topped out at 4.2ish. 266*17=4.5GHz so 16x will probably be as high as you can go.
  2. Well my mobo doesn't support a C2D :( , so it doesn't do me much good to compare them.

    I am running the multiplier at a 16x now based on what I saw from old reviews while doing research last night. Many have stated that going to 16x (=4.267GHz) can be done on a 965 without anything more than stock cooling. That's without changing anything other than the multiplier.

    My mobo (d955xbk) gives me the option of bumping the FSB from 1066 to 1333. Should I try it? It seems like a big jump, but some have said that's alright for a dual-core processor. I can also adjust voltages, but haven't needed to do so yet (nor do i have any experience with it).

    I was thinking of bringing down the multiplier, then bumping the FSB to 1333 and seeing how it goes with some stress tests. If all is stable, then bringing the multiplier up again. Presumably, the PC will drop out on me at some point, at which time I will bump the CPU voltage a little bit. Does that sound like a good plan to get as much as I can from this rig? [Keep in mind I have a liquid cooler for the CPU]

    Any advice is appreciated! Thank you!
  3. I know your board doesn't support C2D CPUs, but if you spent a lot of money (and definitely anything higher then $100.) you were better off dumping your current setup and getting a C2D and new board. I've seen used C2D chips for $40, and I'm sure you can find a good S775 board for $60. (reuse your current DDR2 ram.) Such a setup would greatly outperform that old P4, while not costing you much more.

    Going from memory 4.2GHz-4.5GHz is the top end of what that chip can do. (not counting high end H2O loops or LN2.) However you get there is up to you and your equipment. If you want to do it with just multiplier bumps or FSB increases it doesn't really matter.
  4. I hear where you're coming from in terms of general advice, but right now I'm sitting in front of a computer with a D955xbk mobo and a 965EE chip. I'm just trying to find out how to make the best of what I've got. Cool?

    I've seen some old reviews of this chip, and the predecessor 955. Here's a few that discuss their overclocking achievements.
    ~1.2GHz OC with stock cooling of 965EE
    ~600MHz OC with stock cooling of 955EE
    ~1.3GHz OC with unknown cooling of 955EE

    So it sounds like the chip has more to give, especially since I have water cooler on it. But I've never done any overclocking, so I'm still learning.

    Do you think it's safe to bump the FSB from 1066 to 1333 (and drop the multiplier a bit)? My mobo doesn't let me do small increments. Or should I just keep cranking the multiplier until I lose stability?

  5. Every chip is different, you'll have to try both and see. If you have the multiplier adjustment I'd start with that. If it can handle 17 or 18 that's great. Don't be afraid to bump the Vcore up some however as you'll need more juice the faster it gets.

    I'm not trying to be a jerk either. I know you are setting in front of the computer you have, but maybe its not the computer you should use. You also said "water cooling kit". Some of those kits aren't any better then good air coolers, so don't be totally shocked if/when you get a heat issue. Not saying it will happen, but unless your "kit" is really a "loop", you might strain it with your high TDP chip.

    Like I said, try it both ways and see which gets you farther. And please report back as I'd love to hear how it went. I'm going to guess CPU multiplier will get you farther, but I also bet a small mix of each will be best of all.
  6. Bad news :-(

    I was having no problems at x16, so I went back into bios and tried x17. However, now the computer wont even boot. The video card doesn't even check in, which is usually the first thing I see on the monitor when I turn on the computer.

    I guess that means I have to bump the processor voltage a bit to reach x17 and beyond. Now I gotta figure out how to get back into bios to get it working again.
  7. Well that was a real pain the buttocks. The Intel-recommended method for recovery from a bios-setting-induced non-booting state is to manually reset the bios with the .bio file on either a USB stick or a CD. NEITHER of these did anything at all. No matter what I tried with putting the jumper into "reset" mode (as instructed by Intel), nothing ever showed on the screen.

    Frustrated, I pulled all harddrives and CD drives from the system, still nothing.

    Super-frustrated and about to give up, I put the jumper into the "reset bios passwords" mode... AND IT WORKED! By putting the jumper into this third option, it let me get into the bios where I could change passwords if I wanted, but it also let me reduce the multiplier! Yes!

    Rebooted (replaced jumper and reattached windows harddrive) and got into Windows, albeit VERY slowly. I rebooted again, and it started moving a little faster. I shut down, reinstalled all hardware and rebooted twice more. Now it seems to be back to full speed. Success!

    Well that's all for now, i'm gonna rest on my laurels and watch some Fringe with a beer. I think I deserved it. The saga will continue tomorrow.
  8. Sucks. OCing can be a PITA at times.

    Tomorrow you're testing the FSB changes?
  9. I'm not going to mess with the FSB for now, I've decided. I'm just going to try to work up the multiplier and the CPU voltage. Let's keep it simple, one variable at a time. Once I decide on a max CPU speed, maybe I'll go back and mess with FSB... maybe.

    At any rate, I'm back at 16x multiplier, CPU running at 4.267GHz. I want to go further, but I think going to 17x without bumping voltage will just fail to boot like last time. So I think I have to raise the CPU voltage. The default is 1.3250. I've read that max safe is 1.4, but it's best to go up in small increments.

    So, first I bumped up the CPU voltage from 1.3250 to 1.3375, but kept the 16x multiplier. When I boot into windows, I see no discernable difference in my reported voltages. While sitting at desktop, Speedfan is reporting anywhere from 1.20-1.24, which is the same range I always see (tested at 14x, 15x, and 16x, all at default 1.3250 bios setting).

    Any idea how much I have to bump the bios voltage setting before I can safely jump to 17x? Is there any way to establish better feedback on voltage settings without reaching a point where I have to reset jumpers to recover?
  10. If you don't see your voltages matching, turn off speed step. You can always turn it back on once you get things settled. On a similar note, as long as cooling isn't an issue just set your Vcore at max. Again, you can always turn it back down once you get things settled. (if not max, then whatever max setting you are comfortable with.)

    So I would set Vcore to 1.4V, disable speed step, and try for 17x again. If it works then great/try 18x. If it fails then there isn't much you can do. You could set it back to 16x and start upping the FSB to see if you can hit 4.3 or 4.4GHz. Remember then when you start changing the FSB you'll need to worry about your ram settings as well.
  11. Update:
    I got comfortable with 1.3375 and bumped it again to 1.3500. System booted up just fine, and it was acting normal. Temperatures were a little higher, but still never got over 65.

    However, then I turned on Prime95. After less than 10 minutes, I got a BSOD. System reboot, I backed it down to 1.3375. Re-ran Prime95, and got same BSOD. Now I'm back to stock 1.3250, with a 16x multiplier.

    At no point in this did I ever see the temperatures break 70. Temps spiked as soon as Prime95 started, but then dropped and stabilized close to 62. Then, several minutes later and seemingly out of nowhere, BSOD.

    So it seems raising the CPU voltage is a problem.
Ask a new question

Read More

Intel Pentium Overclocking