Multiplier / Fsb Problems

[FalconsPrey]

Hello again..

 

I have another problem with OCing my old AMD Athlon 64 4000+.. Rigth now I have it stable at 2808 Mhz with 234 x 12, and Vcore varying from 1,49 to 1,52. My goal is to reach 3Ghz, but when I clock it further then 234 x12 it becomes unstable, and increasing the voltage above 1,52 is too risky for my taste.

I read that I should decrease my multiplier and increase my FSB, and so I tried. I set it to 280 x 10 and the thing wouldnt even get past the detecting the drives screen, saying the system gone into safe mode and I should change the CPU clock. I dont get why it says that. When i changed it to 254 x11 it managed to start Windows to the logon screen, but there it was giving me one error message after another. When I reset it to 234 x12 chkdsk came up and found a load of errors. I dont understand why that would be. If it helps, I have a Biostar NF4 Ultra A9A Mainboard. Temperature is also in optimal levels.

I'd appreciate the help a lot.

 

FalconsPrey

Edited January 16, 2009 by FalconsPrey

[Zertz]

Make sure you drop the HT multiplier to 3x-4x otherwise the HT link will work at FSB * 5 so that would be 1400 MHz in your case. Try to keep this around 800-1000 MHz

[FalconsPrey]

Okay... that worked. Now the thing is that if I OC it further, the system becomes unstable. Now I have it at 285 x10 with a HT multiplier at x3. I ran Prime95, and done the Small FFT Test, and it failed after 9 minutes at the 4th 8K test.

If it already fails here, I dont know what to do to get it to 3Ghz and run stable, as the Vcore is already high. On here I read that people managed to get the same CPU, with the same core, to 3Ghz, using lower voltages then I'm running mine at.

Sorry, I'm a newbie.. Could u plz explain to me how to get it to 3ghz stable?

[DLS2008]

Just because one person did it, doesn't mean you can. Not all CPUs are created equal. Some are better than others. Correct me if I'm wrong, but I believe all CPUs are made the same (well, at least in there range, 45nm duals are one, 45nm quads another, etc.) and they are labeled based on their performance. So a Q9550 is the same as a Q9450, it just did better on Intel's tests. Within the same line of CPUs, there can be separate steppings and some are better than others (an E8400 with C1 stepping isn't as good at OC'ing as an E8400 with E0 stepping). Within the same stepping, there are different batches. Some batches are better than others. You can find hardcore OC'ers listing the batch numbers with their OC settings so people can look for other CPUs from the same batch. On top of all this, there are limits that your motherboard has and can't pass. Some motherboard chipsets can run faster than others and even with the same chipsets, some manufacturers are just faster. There are many factors to consider. Your RAM also has an effect and most likely you're not running exactly the same setup that the people who got it to 3GHz were. You may have simply hit the limit for you hardware setup.

[Kuronin]

Indeed, drop your DRAM speed to a lower setting, If you are running at 800MHz, you may want to drop down to 667 or 533. When you increased your FSB frequency, you sped up your DRAM as well as your processor. That should make it stable.

 

And yes, you may be at the end of your rope, I was hoping for 3.0 on my 9550, but I only got a pitiful 660MHz boost at 2.86GHz.

[FalconsPrey]

Yeah, seems I am at my limit.. Lowering the DRAM speed did nothing but destablilize my system even more. Pity. Seems like i just get a pitiful 425 Mhz boost from 2,4 Ghz.. But well.. better then nothing. Thanks.

[Kuronin]

Do not concern yourself with that, it may be that your board is not a good overclocking board. I can only get 2.7 on my 4600+ (300MHz boost) and 2.86 with my 9550 (680MHz boost)

Simply put, AMD chips are not much for overclocking.