CheckEmoji Community · the emoji forum
🏠 Home 🆕 What's new ❓ Unanswered 🔥 Popular 📡 RSS Members 👥 0 online log in · register
Home › IT › Hardware › Desktops › Overclocking (Main Thread)

Overclocking (Main Thread)

Started by dustypilot23 · · 👁 236 views · 3.3K replies

📡 Subscribe to replies

Participants dustypilot23Morgan Ortiz69John Palmer4driftingnomad68neonhawk53rustyrider40brightgull15boldhound8Bradley Myers2driftingmarlin34Aelectriceagle40placidrangeramberorca192Rebecca Parkerwiredjackal7goldenowl73Jose Cook17Tyler Lee57Alex Palmer66Grace Palmer39Scott Morris8briskheron10Thomas Ward9 …
Megan Ruiz13 Megan Ruiz13 Active Member
79 messages
joined Apr 2016
#2321 ·
Edward Peterson70 said:It depends. If the CPU is unstable at that frequency, you just bump up the voltage.

Same goes for memory—if it’s acting up at those speeds, increase the voltage.

How do you actually spot instability in a CPU or RAM caused by voltage issues?
Basically, what are the telltale signs when a PC starts acting out because it needs more juice?

And how do you find that sweet spot for voltage? Is it just trial and error until things stop crashing?

Sorry, I'm a bit lost here. 😁
Rachel Fox4 Rachel Fox4 Member
12 messages
joined Apr 2009
#2322 ·
Gary Anderson38 said:How can you tell if CPU or memory instability is being caused by power delivery issues?
Basically, what are the most common symptoms when this happens, and does it mean I need to bump up the voltage?

And how do you actually find that "sweet spot" for voltage? Is it just a guessing game until you hit it?

Sorry, my typing is a bit messy here 😁

High temperatures, insufficient current, macOS crashing, or seeing errors pop up during stability testing software...

You simply increase the voltage as long as the component remains unstable. If it isn't stable, you add more juice. It’s really that simple.
Megan Ruiz13 Megan Ruiz13 Active Member
79 messages
joined Apr 2016
#2323 ·
Fair enough.

One more thing regarding voltage. Is it actually okay to push past the manufacturer's rated limits for a CPU?

Or am I just asking for a dead processor every single time?

When you guys overclock, have you ever actually found chips that could handle going beyond those factory-recommended specs?
Rachel Fox4 Rachel Fox4 Member
12 messages
joined Apr 2009
#2324 ·
Gary Anderson38 said:I understand.

I have another question regarding voltage. Is it permissible to exceed the maximum voltage limits specified by the processor manufacturer?

Is there a guarantee that a processor will definitely die if handled that way?

When you all approach overclocking, do you ever find yourselves discovering processors that can actually handle pushing past those factory-recommended limits?

Personally, if I were running a Core 2 Duo 24/7, I wouldn't push the voltage past 1.40V—provided, of course, that the cooling setup is solid.
For the Core 2 Quad, I’d suggest setting it at around 1.45V.

You should probably spend some time lurking on the extreme overclocking forums. Some of those enthusiasts tend to push their voltages quite high.
Most of those folks are just doing it for a quick stress test or some CPU-Z validation. Usually, those processors aren't actually stable at those speeds; everyone just wants to squeeze out the absolute maximum frequency possible for the benchmark. If they were aiming for actual stability, the clock speed would be noticeably lower.
Megan Ruiz13 Megan Ruiz13 Active Member
79 messages
joined Apr 2016
#2325 ·
I’ve been lurking on this forum quite a bit lately.

http://forums.extremeoverclocking.com/t261409.html

I was curious to hear about your overclocking experiences—get some actual details and see what's what.🙂

I might try my hand at OC myself soon. For now, I'm just reading, asking questions, and learning the ropes. 😁

Knowing me, though, I'll probably end up breaking something or frying a component. 🤣
Rachel Fox4 Rachel Fox4 Member
12 messages
joined Apr 2009
#2326 ·
Just stick to the sequence and take your time; if you follow the steps, nothing should go wrong.
👍
Besides, there is a massive amount of utility software available these days specifically designed to prevent you from accidentally damaging or frying a component.
Megan Ruiz13 Megan Ruiz13 Active Member
79 messages
joined Apr 2016
#2327 ·
Could you walk me through this with a specific example? Here’s what I'm seeing on my end...

Everest shows my CPU is overclocked by 2%. It says the default CPU FSB should be 266, but mine is sitting at 272, even though I haven't touched a single setting.

I'll post two screenshots so you can see. When I check the motherboard info, the FSB is listed at 272 MHz, which matches what I'm seeing. I do have Intel SpeedStep enabled in the BIOS...

Now I've got another question. Since that's the case, both Everest and SiSSandra are reporting my memory frequency at 408 MHz. By default, it should be 400 MHz—meaning an effective speed of 800, but I'm getting 816.

Is it actually possible for that tiny 2% OC to affect the memory speed like that? Looking at the timings, 400 MHz should be 5-5-5-18, but mine is currently 5-5-5-15. Does that mean the motherboard automatically adjusted the timings? And if so, why would it bother for such a negligible 2% increase, assuming it's even an overclock at all?🤷

Maybe this is just how it's supposed to work.😁

Also, what would happen to the RAM if I bumped the FSB up to hit 3 GHz? My multiplier is 9, and the minimum is 6...

Oh, and SiSSandra is throwing a warning at me to lower the FSB because "overclocking is harmful to hardware." I'm looking at it like, WTF? What overclock?🤣

Anyway, here are the screenshots. Can someone explain this? I haven't changed anything.

http://img150.imageshack.us/img150/3078/prochd5.jpg

http://img150.imageshack.us/img150/5034/plocakr8.jpg

This screenshot shows the speed under load. Since SpeedStep is on, it shows 1.6 GHz up top, when it should normally be 2.4. But because of that 272 FSB, I'm actually hitting 2448. Lol, what kind of "increase" is that?😁

I just ran some random benchmark to speed up the CPU, you can see it here...

http://img233.imageshack.us/img233/255/brki4.jpg

2448.4 mhz 🤷

I'm trying to push the clock a bit—nothing crazy, maybe 2.8 or 3 GHz—but I have no idea how the memory will react. I'm hoping the board just handles it automatically...

I was just prepping for this, and then this whole thing caught me off guard.
Harold Lopez31 Harold Lopez31 Active Member
75 messages
joined Jan 2008
#2328 ·
wiredmarlin5 wiredmarlin5 Newcomer
2 messages
joined Apr 2009
#2329 ·
That’s actually pretty sweet🙂 . . . I mean, those E4500 chips weren't exactly legendary when it came to overclocking potential, but hey, if you know what you're doing, they can really take off under the right conditions... 🙂
Rachel Fox4 Rachel Fox4 Member
12 messages
joined Apr 2009
#2330 ·
Finally, someone is actually posting about overclocking here. 🙂
🙂 It’s about time we got some info 😁 regarding
Sam James7 Sam James7 Newcomer
5 messages
joined Apr 2008
#2331 ·
Hey Edward Peterson70, I’ve got a quick question for you now that I’m here—which one actually has a better price-to-performance ratio? Is it the Core 2 Duo E8400 or the Core 2 Quad Q6600?
Rachel Fox4 Rachel Fox4 Member
12 messages
joined Apr 2009
#2332 ·
Harold Lopez31 said:🙂 If only the cooling were any better; the stock cooler is actually worse.

Really?

Just stick the computer in the fridge 🙂

Sam James7 said:Hey Jimmy, I have one more question. Which processor is actually better overall, the C2D E8400 or the quad q6600?

Intel Core 2 Duo E8400
Megan Ruiz13 Megan Ruiz13 Active Member
79 messages
joined Apr 2016
#2333 ·
Harold Lopez31 said:Gotta hop in the shower, be right back.

Can you walk me through what you did with the memory speed? Did you just let it scale with the FSB, and if so, how high could you push it before it lost stability?

Or did you manually dial it back? DRAM Frequency?
Harold Lopez31 Harold Lopez31 Active Member
75 messages
joined Jan 2008
#2334 ·
Validation says☕

Basically, just keep the fsb:dram at 1:1. A high multiplier shouldn't cause any headaches.
Kyle Sanchez4 Kyle Sanchez4 Newcomer
4 messages
joined Dec 2009
#2335 ·
You need to scale back... RAM can only go so far...
George Ross41 George Ross41 Active Member
70 messages
joined Jan 2008
#2336 ·
http://valid.x86-secret.com/show_oc.php?id=353058

The temperature is hovering right around 98-100 degrees Fahrenheit
while the load is sitting steady at about 42-43 percent 😉
Megan Ruiz13 Megan Ruiz13 Active Member
79 messages
joined Apr 2016
#2337 ·
Harold Lopez31 said:It says on the validation page☕

Basically, fsb:dram = 1:1. A high multiplier shouldn't cause issues.

Got it. You bumped the multiplier to drop the FSB. I see you.

Just saw the screenshots. 😁

I can't touch my multiplier; 9x is as high as it goes.

So I'll have to slow down the memory and pick different ratios.

Wait, if it's 800MHz RAM, it runs at 400 in the system, but that means it could theoretically push past 400 closer to that 800 mark, right?
I'm guessing it depends on how good the sticks are.

Logically, should I just throttle it now, or try to see if it can actually handle the full 800?

Hey George Ross41, looks like you bumped your multiplier too, didn't it?

221.22 * 15

And the memory

RAM Speed : 414.8 MHz (CPU/8) @ 5-5-5-18

You've got that same 800MHz PC6400 RAM, too.

Right here http://valid.x86-secret.com/show_oc.php?id=353003
Here it's 330..

Does that mean it's smarter to just avoid overclocking the memory?

I'm going to try pushing the CPU to 2.8 GHz right now. Everything else is set—PCI at 100, disabled C1 SpeedStep. I'll mess with Eupal later; I just want to see if this works first.

The only thing is, if I use "by SPD" for frequency, will increasing the FSB also crank up the memory? Or should I manually set the ratios to keep it at 400 MHz...

I'll probably just lower it to avoid breaking something... then we'll see what happens.

Never overclocked anything in my life, but I'm curious....

That Intel Core 2 Quad Q6600 can hit a clean 3 GHz without much effort... from what I've read. 😁
Rachel Fox4 Rachel Fox4 Member
12 messages
joined Apr 2009
#2338 ·
Go ahead and read that overclocking guide on PC Gamer. And don't try to lie to me about having already gone through it 🙂
wiredmarlin5 wiredmarlin5 Newcomer
2 messages
joined Apr 2009
#2339 ·
Harold Lopez31 said:Thanks! Is there any way to check the processor batch if I don't have the actual box handy?

Well, you can usually just read it right off the chip itself... 👍

@Gary Anderson38/">@@Gary Anderson38, an Intel Core 2 Quad Q6600 should hit 3GHz without breaking a sweat at all...

First things first, you really ought to drop all your voltages down to the absolute minimum, though you'll want to set the CPU voltage to a specific, fixed number instead of letting it wander...

From what I recall, the Q6600 typically runs at 1.25V by default, so maybe try setting it to 1.3V and bump that processor up to 3GHz, while keeping your memory frequency somewhere around 800MHz... and please, do be careful with the memory voltage, because if you push it to 2.1V (+0.3V), it’s going to be a disaster...

If it actually boots up successfully, let us know how the BIOS looks and what those core temperatures are doing... 🙂
Megan Ruiz13 Megan Ruiz13 Active Member
79 messages
joined Apr 2016
#2340 ·
Edward Peterson70 said:Just go read that overclocking guide on Tom's Hardware and stop lying about having read it.🙂

http://forum.tomshardware.com/showthread.php?t=88200
Thanks for the tip, found my answer here.
😁
Alright, heading into the overclocking now. I'll post results if the PC actually survives.

sry everyone if I was being annoying.😛

You must log in or register to reply here.

Log in Register

🔗 Similar threads