help...kind of new to oc'ing cpus

elitebear4 · Jan 11, 2005

im running on the rig below and just a minute ago, i simply pushed my FSB from 204 to around 220, just to experiment. this was around 2000mhz and was stable. even without upping my ram speed to equal levels of my cpus FSB, it was still stable? i thought i read that there would be instability problems if fsb and ram speeds werent equal. can someone explain this?

as i said, im very new to this, so apologies

hitechjb1 · Jan 11, 2005

By default ram setting or manually set to max memclock = 200 MHz, the memory module bus frequency is running in 1:1 ratio with HTT (or FSB in bios).

So when HTT went from 204 to 220 MHz, memory bus frequency would follow.

You PQI 3200 has a good chance to be overclocked to much high, not uncommon to 250 - 280 MHz with timing of 2.5-4-4-8 1T at 2.x V.

hitechjb1 · Jan 11, 2005

Further, from you list of hardware, you have a good chance to achieve 2.5 - 2.7 GHz.

To get to 2500 MHz
HTT = 278 MHz
CPU_multiplier = 9
max memclock = 200 MHz (aka 1:1 ratio)
memory_bus_frequency = 2500 / 9 = 278 MHz
memory would be overclocked to 278 MHz using 2.5-4-4-8 1T 2.8 V

To get to 2700 MHz
HTT = 300 MHz
CPU_multiplier = 9
max memclock = 166 MHz (aka 5:6 ratio)
memory_bus_frequency = 2700 / 11 = 245 MHz
memory would be overclocked to 245 MHz using 2.5-4-4-8 1T 2.8 V

These are just two example operating points, you can explore something in between by adjusting HTT.

hitechjb1 · Jan 11, 2005

In the last two posts, I assumed your PQI PC3200 is the PQI Turbo PC3200 which is based on TCCD memory chips.

If not, please post otherwise, as non-TCCD based memory modules may not be able to run at such high bus frequency.

elitebear4 · Jan 11, 2005

heh heh. im running on cheap 2.5 cas ram

elitebear4 · Jan 11, 2005

so it would be okay to just leave my FSB at 220? would there be any stability issues? im feeling a bit paranoid right now, as i dont want my system to break =/

hitechjb1 · Jan 11, 2005

For stability, it has to be tested. Run some programs such as Super PI, 3D mark 01, Prime95, ... to test it.

At 220 MHz of HTT, you CPU would only be running 220 *9 = 1980 MHz, it is far from getting into unstable region of overclocking.

What exact model is your PQI PC3200.

For your 3000+ Winchester, very good chance to overclock to 2.5 GHz (at least).

If your memory module is PQI turbo 3200, try as suggested in the previous post.

If memory is not PQI turbo (TCCD), try run memory slower as follow:

Try this first:
CPU 2300 MHz
HTT = 256 MHz
CPU_multiplier = 9
max memclock = 166 MHz (aka 5:6 ratio) <---- set this in bios
memory_bus_frequency = 2300 / 11 = 209 MHz
(PC3200 memory should be able to handle this, even non-TCCD)

When it works and you feel comfortable
Then try CPU to 2500 MHz
HTT = 278 MHz
CPU_multiplier = 9
max memclock = 166 MHz (aka 5:6 ratio) <---- set this in bios
memory_bus_frequency = 2500 / 11 = 227 MHz

etc

elitebear4 · Jan 11, 2005

http://www.newegg.com/app/ViewProductDesc.asp?description=20-141-164&depa=0

wait...how would i access memory bus frequency?

hitechjb1 · Jan 11, 2005

In bios, set manually max memclock to 200 MHz (1:1) or 166 MHz (5:6), then memory bus frequency will be locked to the CPU frequency as the HTT frequency is adjusted.

To overclock the A64 3000+,
set AGP to 67 MHz
set Vcore to 1.4 - 1.5 V
set the max memclock to 200 MHz (aka 1:1) or 166 MHz (aka 5:6) (as described above)
set CPU_multiplier to x9 (default)
set LDT to x4 or x3

the remaining is HTT.

By adjusting HTT, the CPU frequency and memory bus frequency will change accordingly.

CPU_frequency = HTT * 9
memory_bus_frequency = CPU_frequency / 9 (1:1 ratio)
memory_bus_frequency = CPU_frequency / 11 (5:6 ratio)
HyperTransport_frequency = HTT * LDT

Try the setting as suggested in my last post.

Tiesto · Jan 11, 2005

hitechjb1 said:
For your 3000+ Winchester, very good chance to overclock to 2.5 GHz (at least).

Try this first:
CPU 2300 MHz
HTT = 256 MHz
CPU_multiplier = 9
max memclock = 166 MHz (aka 5:6 ratio) <---- set this in bios
memory_bus_frequency = 2300 / 11 = 209 MHz
(PC3200 memory should be able to handle this, even non-TCCD)

When it works and you feel comfortable
Then try CPU to 2500 MHz
HTT = 278 MHz
CPU_multiplier = 9
max memclock = 166 MHz (aka 5:6 ratio) <---- set this in bios
memory_bus_frequency = 2500 / 11 = 227 MHz

sorry to hijack this thread, but thanks for your easy-to-follow settings. :clap:

I'm running now stable with the first set of settings you posted (I also have a 3000 winchester), and by the weekend, i'll try to shoot for 2.5 Ghz.

Damn, it should be a *crime* not to overclock a 3000+ winney. And best of all, I've got mine running at 2.3 Ghz easily (have to run 5/6 ratio, i dont trust my ram yet; need to do memtest on it first), stock 1.4V, and a 350W PSU (17A on 12V rail), and 12V line holding steady at 11.95V. And I definately recommend Neo2 Platinum board. It seems to be able to run 300 FSB standing on its head. I cranked mine up to 255 and it went up 1deg C.

hitechjb1 · Jan 11, 2005

Tiesto said:
sorry to hijack this thread, but thanks for your easy-to-follow settings.

I'm running now stable with the first set of settings you posted (I also have a 3000 winchester), and by the weekend, i'll try to shoot for 2.5 Ghz.

Damn, it should be a *crime* not to overclock a 3000+ winney. And best of all, I've got mine running at 2.3 Ghz easily (have to run 5/6 ratio, i dont trust my ram yet; need to do memtest on it first), stock 1.4V, and a 350W PSU (17A on 12V rail), and 12V line holding steady at 11.95V. And I definately recommend Neo2 Platinum board. It seems to be able to run 300 FSB standing on its head. I cranked mine up to 255 and it went up 1deg C.

Glad to hear it works for you.
Even you think your memory is slow, you may try this now to bring the 3000+ higher.

Overclocking 3000+ Winchester with slow PC3200 memory

CPU 2500 MHz
HTT = 278 MHz
CPU_multiplier = 9
max memclock = 133 MHz (aka 2:3 ratio) <---- set this in bios
memory_bus_frequency = 2500 / 14 = 179 MHz

CPU 2600 MHz
HTT = 289 MHz
CPU_multiplier = 9
max memclock = 133 MHz (aka 2:3 ratio) <---- set this in bios
memory_bus_frequency = 2600 / 14 = 186 MHz

CPU 2700 MHz
HTT = 300 MHz
CPU_multiplier = 9
max memclock = 133 MHz (aka 2:3 ratio) <---- set this in bios
memory_bus_frequency = 2700 / 14 = 193 MHz

All three settings can bring CPU to 2500 - 2700 MHz even using slow PC3200 memory. Try it, don't wait.

Overclocking setting for various bus frequencies (post 8)

Some overclocking scenarios for 939 Winchester

elitebear4 · Jan 11, 2005

alrighty! thanks a million, hitech!

anyways, would setting the memory at slower speeds, still not slow down your performance for some reason?

hitechjb1 · Jan 11, 2005

elitebear4 said:
alrighty! thanks a million, hitech!

anyways, would setting the memory at slower speeds, still not slow down your performance for some reason?

Case 1:
If both CPU and memory are of fast type (e.g. TCCD), can one can overclock both to high level.

Case 2:
But if CPU is fast like Winchester to 2.5 - 2.7 GHz on air, but memory is slow, then one can slow down the memory to whatever it can take and still be able to overclock the CPU to 2.5 - 2.7 GHz.

The overall performance of Case 2 will be worse than that of Case 1, since memory is slower. But until better memory is used, Case 2 is the best option for that given hardware.

Tiesto · Jan 11, 2005

Alright, i'll go crank her up even higher.

Just played Half Life 2 (1024rez, all details set to High, and it was silky smooth... *BIG GRIN*) for about half hour, and CPU temp held steady at 40C. I think she's got a lot potential in her.

@ elitebear: i see what you're saying, but a CPU running at 2.7Ghz with 200mhz memory is still faster than a CPU running at 2.3Ghz with 200mhz memory, right?

although i'm confident i could bring them up to about 230+, but i want to use memtest on them at those speeds. right now, i just wanna crank this 3000+ til it bleeds.

@hitechjb: do most winchesters need their voltage cranked up past 1.4v when getting into the 2.6 / 2.7 Ghz range? I'd rather stick to 1.4V (or even less) until i get a better PSU. Don't want to overload my 350W.

hitechjb1 · Jan 11, 2005

Tiesto said:
Alright, i'll go crank her up even higher.

Just played Half Life 2 (1024rez, all details set to High, and it was silky smooth... *BIG GRIN*) for about half hour, and CPU temp held steady at 40C. I think she's got a lot potential in her.

@ elitebear: i see what you're saying, but a CPU running at 2.7Ghz with 200mhz memory is still faster than a CPU running at 2.3Ghz with 200mhz memory, right? although i'm confident i could bring them up to about 230+, but i want to use memtest on them at those speeds. right now, i just wanna crank this 3000+ til it bleeds.

@hitechjb: do most winchesters need their voltage cranked up past 1.4v when getting into the 2.6 / 2.7 Ghz range? I'd rather stick to 1.4V (or even less) until i get a better PSU. Don't want to overload my 350W.

From some results I've seen on Winchester, 1.4 V can roughly get it to 2.5 GHz. To get to 2.6 - 2.7 GHz, 1.5 - 1.55 V would be needed in general.

hitechjb1 · Jan 11, 2005

As far as what addition power and current is needed,

going from 1.4 V 2.5 GHz to 1.55 V 2.7 GHz, the increase in power and current on the 12 V is about 32%.

elitebear4 · Jan 11, 2005

i havent tried this yet, as im off limits to my computer (defragging it), but i was wondering if i would experience any issues keeping perfect stability, due to temp issues. my antec superlanboy runs extremely hot and my CPU is at 40 C to about 37 C at best, with stock settings on IDLE. assuming that my idle temp rises around 10 C more, would i still be in good shape?

edit: a somewhat off topic question: does oc'ing, without changing voltage levels, and keeping decent temps, still cut the lifespan of your hardware?

hitechjb1 · Jan 11, 2005

elitebear4 said:
i havent tried this yet, as im off limits to my computer (defragging it), but i was wondering if i would experience any issues keeping perfect stability, due to temp issues. my antec superlanboy runs extremely hot and my CPU is at 40 C to about 37 C at best, with stock settings on IDLE. assuming that my idle temp rises around 10 C more, would i still be in good shape?

You are correct about temperature rise of about 10 C going from idle (light load) to full load (such as Prime). For general usage, temperature rise would be less.

With load temperature rise, at 47-50 C, it should be OK for the CPU. You may look at improving case cooling to bring the system ambient lower.

As for any overclocking, experiments, there are risks involved, in return we would get the learning, experience, skill and a faster computer, ....

Tiesto · Jan 11, 2005

thanks for your help, hitechjb

got mine steady now at 2.4Ghz, 1.4V, 267HTT, 5:6 divider (memory at 219mhz), and temps still at 40C even on load, stock HSF + extra Case Fan.

I tried to drop memory to 133mhz in Bios and crank HTT up to 275, but didnt POST. me thinks i'd have to touch my Vcore which i dont want to at the moment, so i'll just have to be happy running my 3000+ at 3500+ speeds 'til i get a new PSU.

gonna go burn it in a bit overnight and see how temp and stability hold.

P.s. my neo2 platinium is set to 4x 267 = 1068mhz. is that too much for the MSi board? would you recommend 3x instead, to ease off on my northbridge?

hitechjb1 · Jan 11, 2005

elitebear4 said:
edit: a somewhat off topic question: does oc'ing, without changing voltage levels, and keeping decent temps, still cut the lifespan of your hardware?

Even if voltage is kept constant, and if one is able to clock the CPU frequency higher, more power is dissipated since active power of a chip is proportional to frequency (at constant voltage), hence the temperature of the chip will increase. Since a main factor that determines life expectancy is electromigration which is a function of temperature, hence life expectancy would also be shortened. The question is by how much putting the chip usage in context, e.g. an overclocker's computer, a mission-critical server, a computer for space exploration in a space-ship, ....

hitechjb1 said:
On CPU life expectancy and tradeoff

There is NO clear Yes or No answer to predict "life expectancy" regarding to a particular CPU based on voltage and temperature. For a given CPU type, they follow certain statistical behavior.

The inverse relationship between frequency and temperature will naturally determine the max voltage and frequency for a given CPU and cooling setup. If overclocking is done properly, such voltage and temperature should be below the max absolute temperature and voltage of a given CPU specifcation (at least true for Tbred/Barton).

For example, assuming nominal voltage is 1.5 V.

Running it constantly at 1.8 V is 20% over nominal and running it at 1.95 V is 30% over nominal voltage.

From electromigration analysis, keeping temperature roughly constant (by cooling), overvoltage from 20% to 30% decreases CPU failure time (life expectancy) by about 10% (failure time reduced from 69% to 59%)

For Tbred B/Barton/Mobile Barton, the frequency gain between 1.8 V to 1.95 V is about 75-100 MHz at 2300-2500 MHz level, it is stating to operate in the diminishing return regime, getting only 75-100 MHz from 150 mV Vcore increase. The gain in overclocking frequency is about 3 - 4%.

The most effective, cost effective overclocking voltage for Tbred/Barton is between 1.5 - 1.9 V, beyond which overclocking is very costly in terms of power supply and cooling with diminishing frequency gain.

So it is a tradefoff between frequency, voltage and life expectancy.

Going from 1.8 V to 1.95 V, one would get 3 - 4% increase in overclocking frequency at the expense of an additional 10% reduction in CPU life expectancy (statisitically).

But then a different person uses a CPU differently and have different objective, some expect to use it for 6 months, 1 year, some for 3 year, ..., some want to squeeze the last MHz for competition and satifaction, so one has to make his/her own judgement and tradeoff, ...

This post discusses this subject in details:

CPU voltage: from stock to max absolute, from efficient overclocking to diminishing return (page 19)

Effect of high Vcore and electromigration on CPU failure time (page 15)
Effect of high Vcore and electromigration on expected failure time for Tbred B/Barton (page 15)
What could damage a chip/CPU permanently? (page 15)
What is gate break-down voltage (page 16)

help...kind of new to oc'ing cpus

Member

Senior Member

Senior Member

Senior Member

Member

Member

Senior Member

Member

Senior Member

Member

Senior Member

Member

Senior Member

Member

Senior Member

Senior Member

Member

Senior Member

Member

Senior Member

Similar threads