Overclocking (Main Thread)

darkowl72 said:My e2200 was running at 280x11, which clocks in at 3.08GHz, with temperatures sitting at 76°C for TAT and 72°C during Orthos. I couldn't push it any further without messing with the voltage, and since what I had was perfectly sufficient, I didn't see much point in forcing the issue...

Now, I recall stumbling upon some discussion regarding undervolting—not in the sense of dropping the frequency just to cool things down or quiet the fans, but rather attempting to lower the CPU's clock voltage without sacrificing actual performance. Since I found myself with a bit of free time today, I figured I might as well see if it was worth the effort...

Working with an 11 multiplier yielded absolutely nothing... other than the realization that this processor is remarkably resistant to undervolting... 😉 At the default 1.45V, CPU-Z reports 1.45V and Drake 1.344V; at 1.43125V, it shows 1.392V and Drake 1.360V; then at 1.425V, it looks exactly the same (!)... and everything remains stable at 280x11 while the temperatures don't budge one bit... (?) It’s all a bit peculiar, but I kept going...

So, I decided to start playing around with the multiplier instead...

(A brief tangent: I previously established that cranking up the CPU voltage while keeping the multiplier at 11 doesn't allow the FSB to move even a single MHz past 280!)

Well, after a fair amount of tinkering and testing, I didn't manage to break anything 😉 and finally stumbled upon a winning combination:

410x8 = 3.28GHz, FSB = 1640
Voltages: BIOS is 1.44375V, CPU-Z shows 1.408V, Drake is 1.392 / 1.376V (mostly staying at the latter, though 1.392 pops up occasionally)
SPD = 2.0, RAM at 410 (1:1), latencies at 5-6-6-19 2T (Auto)

It’s holding steady under Orthos; it’s been over three hours now...

Temperatures

TAT at 66/66 and hasn't crossed a single degree higher
CoreTemp at 68/68
HWMonitor: tmpin0 48, tmpin1 63, core#0 68, core#1 68, GPU core 59, HDD0 45, HDD1 36

The cooler is a Thermaltake Blue Orb II, spinning happily at 1500 RPM (795 idle)
Motherboard is a P35-DS3, memory is 2x 2GB Kingmax 800, and the GPU is a Gainward Bliss 8800 GTS 320

As for why this dropped my load temperatures by a full 10 degrees (!), I honestly can't say whether it's due to the higher FSB paired with a lower multiplier, or perhaps a slight undervolt inherent to this specific configuration... I haven't the slightest clue, but the fact remains that the temperatures are much more reasonable now...

I should also point out the drive temperatures from startup compared to after about eight hours of runtime, even without heavy load:

Seagate BarraCuda 320, started at 30, now 45
Samsung T166 400, started at 24, now 36

Oh wow, 45 degrees for those drives?? What on earth are you doing to them? 😲

Well, nothing 😉 Seagate BarraCuda is what this is... plus it's loud... really loud...

Is it actually possible for the default voltage to be sitting at 1.45V🤷? That feels way too high to me. Also, when you're talking (or rather, when you all are talking—sorry if I'm being too informal here👍 ) about voltages, my understanding is that Drake (at least as far as I'm concerned) isn't the actual voltage under load, but the difference between idle and load. For example:

At 1.43125V, it shows 1.392V and a Drake of 1.360V -> Drake = 1.392V - 1.360V

Don't use the plural with me, I'm flying solo here... 😉

Yeah, it's definitely 1.45... that's how it's been set since I slotted it into the motherboard... and as for that > stuff... I just didn't feel like doing the math myself. I figured if you're smart enough to follow along, you'll get it... 😉

darkowl72 said:The readout says 1.43125V, but the actual reading is stuck at 1.392V. Drake is sitting at 1.360V.

It really should be set to 1.36V in the load line. You’d see the Drake drop from 1.392V down to 1.360V—but honestly, that’s splitting hairs. It doesn't really matter.👍

@darkowl72/">@@darkowl72, look, you’ve got some decent hardware here, but honestly, that cooler you're running is absolute garbage...

1. Just completely disregard whatever voltage the BIOS is reporting, because that doesn't tell the whole story; you really need to be looking at the actual voltage within Windows and specifically when the system is under a heavy load. That gap you see between them? That’s just the Drake effect... In my humble opinion, seeing temperatures climb north of 60 degrees Celsius during a load test is just totally unacceptable... so if I were in your shoes, I’d start by dialing back the voltage a bit, and maybe nudge the frequency down too, if you absolutely have to. I’d probably aim for somewhere around 1.30-1.32V while idling in Windows and then just hunt for the highest stable clock speed you can get at that specific voltage. To be perfectly honest, I don't quite grasp how the multiplier is driving those temps up so high, but 11X feels like way too much... maybe try dropping it to 9x or 10x... whichever one makes you smile more, purely for the memes... 🙂

2. Make sure you have all those helpful settings enabled in your BIOS, like the ones meant to help throttle the voltage, adjust the multiplier, and so on... 🙂

3. If you aren't afraid to spend a little bit of your precious free time, you should just sell that current cooler and pick up a ThermalTake Typhoon... then you'll finally see those temperatures dipping comfortably below 60 degrees...😁

Hey, thanks...

I'm gonna go try this right now. My son jumped on my other PC, so I've been stuck wrestling with these temperatures all day. Turns out my CPU is absolutely cooking at 1.45V without me even realizing it. Dammit, I honestly thought the BIOS was just setting it to default voltage, same as the FSB and multiplier...

I honestly can't wrap my head around why they'd set those default settings to 1.45V in the first place... I mean, really...?!

Intel states that the operating voltage for your processor should be between 1.162V and 1.312V, and if you look at the Thermal Specification, they list 73.3°C as the limit. 🙂

http://processorfinder.intel.com/det...x?sspec=sla8x#

So, yeah, just go ahead and forget about that 60-degree mark I mentioned earlier... staying under 70 °C is perfectly fine... 🙂

Yeah, sure... I took a look at that too... honestly, it didn't seem worth the headache. I flashed the F6 BIOS, which recognized the E2200 just fine. Set the FSB to 200, multiplier to 11, kept everything else at default and locked... I wasn't about to mess around with CPU voltage there.

And look, this piece of junk is running perfectly fine with everything enabled; zero issues so far.

I spent all last night and this morning tinkering with the rig again, and finally, everything is rock solid at 400x8 = 3.2GHz. FSB is 1600, SPD is 2.0, so the memory is sitting at 800. I set the voltage to 1.4000V in the BIOS, but CPU-Z reports it at 1.36V...

In Prime95, the voltage drops to 1.328V, and Ortho shows 1.344V. Temps aren't even hitting 66C...

It’s not some massive overclock, but jumping from 2.2GHz stock to 3.2GHz? Not bad at all.😉

I could probably push it to 420x8, but I'd have to crank the voltage, and then temps would likely go nuts again. Since it can handle up to 73.3C, and with Prime95 running full tilt it stays under 66C, I think I'm in the clear and the voltages aren't too high.

With BIOS at 1.38125, CPU-Z shows 1.344, Prime95 hits 1.312, and Ortho sits at 1.328—Prime95 works, but Ortho crashes after about ten minutes...

And that's using the TT Blue Orb II... haha... yeah, I know it's garbage, but I got it in a bundle and frankly, I can't be bothered to swap it out for something else... though, being honest, I bet the CPU would actually stay cooler with the stock cooler.😉... But the thing I like about the Blue is the massive fan; even if it doesn't cool the CPU that great, it creates a damn windstorm that you can feel throughout the whole case... and that's just idling at 900rpm. At 1500rpm, it feels like a gale in San Diego...😉 and surprisingly, it's actually pretty quiet. When I think back to those Zalman cases (I think it was the 8500), man, those things sounded like lawnmowers under full load.

Anyway, Ortho has been running for over 4 hours now and it's staying steady at 63-64C.

http://www.imagesforme.com/out.php/t32904_vcore3.png

Man, that’s clean. 👍

hahaha... you actually pulled it off? respect...

I've got a bit of a headache here: 🤣🤣
I can't get my CPU overclocked right. Like, sometimes I manage to bump it up to 2.7 GHz (even with the voltage cranked up—it's sitting at 1.350V right now), but that's super hit or miss and even then it acts buggy!! I have no clue if the motherboard is failing me or if it's just the processor?? (I'm leaning towards it not being the CPU since it runs perfectly fine on stock settings, actually better than when I try to OC it!!)😕

Here's what I'm running:
CPU: AMD Athlon 2x 4800+ 2.5 GHz 1mb cache + Thermaltake Blue Orb FX
MBO: MSI K9N Ultra nForce 560
GPU: XFX GeForce 7600GT
RAM: Crucial Ballistix 2x512mb 800MHz & 2x1GB 800MHz
PSU: Western Digital 500 w HDD
HDD: Seagate 250 GB

Any advice or thoughts would be huge!!
And please, for the love of god, don't just trash my entire build!!
THANKS👍

slyheron13 said:I've run into a bit of an issue: 🤣🤣
I'm having trouble overclocking my CPU properly. Sometimes I can manage to push it up to 2.7 GHz (even with voltage adjustments—I'm currently at 1.350V), but that's rare, and even then, it doesn't perform reliably. I'm not sure if the motherboard is the culprit or if it's the processor itself. (I don't think it's the CPU since it runs perfectly fine at stock settings, actually performing better than when I try to overclock it!)😕

My specs:
CPU: AMD Athlon 2x 4800+ 2.5 GHz 1mb cache + Thermaltake Blue Orb FX
MB: K9N Ultra nForce 560
GPU: XFX GeForce 7600GT
RAM: Crucial Ballistix 2x512mb 800MHz & 2x1Gb 800MHz
PSU: Western Digital 500 w HDD
HDD: Seagate 250 GB

Any advice or feedback would be greatly appreciated!!
Please, let's keep the criticism to a minimum!!
THANKS👍

What frequency are you running your memory at?
Did you lock the PCI bus to 100 MHz?
Generally speaking, you should set everything to manual settings.

Edward Peterson70 said:What voltage are you running on the memory?
Did you lock the PCI at 100 MHz?
Just set everything to manual if I were you.

My PCI is sitting right at the standard 100 MHz!
And I’ve got the memory clocked at 800 MHz at 2.00V!
I've heard this K9N Ultra board is absolute garbage!

I somehow managed to overclock my CPU to 2.8, but I'm hitting a wall and can't go any higher. I haven't touched anything except bumping the voltage up to 2.10V, nothing else!!
So, if anyone actually knows why I can't push this processor any further, let me know!

slyheron13 said:I managed to overclock my processor to 2.8 GHz somehow, but I can't push it any further. I haven't touched the memory settings at all, other than bumping the voltage to 2.10V.
If anyone knows why I can't get a higher overclock on the CPU, please let me know!

Try pulling out one of the sticks (the 512MB one), so you're just running 2x1GB, then attempt the OC.

If you have all four slots filled, do you think things might start stuttering once you push the overclock?

Daniel Nguyen3 said:Having all four slots occupied might cause some stuttering during overclocking...

especially when dealing with different memory modules.

Edward Peterson70 said:Pull the memory (the 2x 512MB sticks), leave just the 2x 1GB ones, and then try to OC.

I already tried that and saw absolutely zero difference!!

Now I'm gonna mess around with ntune and see what kind of nonsense it spits out!