Overclocking (Main Thread)

Gary Hernandez26 said:Sensors don't lie; they just aren't calibrated properly. This isn't rare, and you can usually fix it by updating the BIOS.

😬 Typical ignorance. That's just not how this works. If they aren't calibrated correctly, then they are lying, right?
A single BIOS update covers an entire line of motherboards, yet individual boards from the same series can report different temperatures. You could have two identical CPUs where one shows accurate readings and the other shows 15 degrees higher. No BIOS update is going to solve that discrepancy for me.

neonmarlin27 said:😬 This is just a classic case of misunderstanding things. It doesn't mean what you think it means. If they aren't calibrated correctly, it just means they're inaccurate—it doesn't mean they're lying, right?

If you actually understood what the word "lying" implies, you wouldn't be making such nonsensical claims. Something can be malfunctioning or incorrect without it being an intentional lie.

neonmarlin27 said:😬
A new BIOS update is generally designed for all motherboards within the same line.

Well, why don't you try downloading a BIOS from a different board of the same brand and see if you can install it?

neonmarlin27 said:😬
Every motherboard from the same series might report temperatures differently. One could show the correct temp, while another shows 15 degrees higher using the exact same processor. There is no way you could fix that kind of discrepancy with a single BIOS update.

The issue lies in how the temperature is being read, not necessarily with the sensor itself. If you had ever encountered this specific problem and actually solved it, you would know the answer.
I'm certainly no electrical engineer, and I don't claim to understand the intricate circuitry of a motherboard, but having worked in a repair shop, I know it isn't rare for a board to report incorrect temperatures. When people hear arguments like yours, they end up thinking there's no solution at all, which leads them to waste money buying $100 coolers unnecessarily.
I have no desire to get into a back-and-forth argument; I was simply trying to offer some useful advice.

Gary Hernandez26 said:Just so you know, saying it "lies" is nonsense. It can be malfunctioning, but it can't lie!!!

Try downloading a BIOS from a different board of the same brand and attempt to install it.

The issue lies in how the temperature is read, not the sensor itself. If you had ever encountered this specific problem and solved it, you would know the answer.
I am not an electrical engineer and I don't understand motherboard architecture, but I have worked in repair shops. It is not uncommon for a motherboard to report incorrect temperatures, and because of
statements like yours, people think there is no solution and waste money buying Zalman coolers from $100 for no reason.
I have no interest in arguing; I just wanted to offer useful advice.

Good grief! I was referring to the BIOS on identical motherboard models. When I say it "lies," I believe my meaning is clear.

Regarding the technical issue: if it is a reading error, then the fault lies either with the sensor or with inaccurate calibration of the reading scale. In mass production, it is impossible to verify that every single sensor is perfectly accurate.
Don't assume I haven't dealt with this. I certainly have, on my old Pentium 4 motherboard. The only solution is to measure the temperature with precise instruments, determine the deviation from what the BIOS displays, and then mod the BIOS so the readings are calibrated accordingly.
The other option is installing a new sensor and a separate panel on the case to show the true temperature, but both options are far too complicated for 99.99% of users. If you have another solution, feel free to share, but in my experience—and I'm not bragging, I've been working in this field for five years since middle school—these are the only ways.

As for your comment about wasting money on $50 coolers (if it really is a Zalman, you might be right), I find it hard to believe a technician would say something like that.
First off, tell me which stock cooler could handle a Prescott overclocked by 40-50% with increased voltages, not to mention the noise levels involved. I have years of experience with this, so you don't need to explain it to me, even though you likely have experience yourself as a technician.

Personally, I have never suggested buying a new cooler if we are talking about a CPU at stock settings. I know exactly how the temps run on certain
processors with stock cooling, which allows me to recognize when a motherboard is "lying."

In the end, I have no desire to argue either.👍 Everything I am writing is based on personal experience, not reading garbage about "bugs" and such.

I believe I have explained the problem clearly now, as very little can be gathered from your posts. I admit my previous post was rushed and poorly written, but you haven't provided anything concrete either.😉

Moderators, my apologies for this massive off-topic tangent, but I had to write it.🙂

Sam Ward53 said:You could probably push it up to 3.8, provided you upgrade the RAM and swap in a higher-quality CPU cooler, like a Zalman.

Right, the power supply. I totally missed mentioning how important a solid PSU is.

Forget 3.8 GHz. You won't get stability with a Zalman7000 or any other air cooler here.
The limit for an E6400 on air is 3.6-3.7 GHz at 1.55V. You’ll need a lot of luck and high-end memory to hit that. Anything higher requires water cooling.
Any decent power supply rated at 350W or more will do the job.

Got a question: How much can I actually push my 2.53GHz Celeron? I'm running an ASUS P5W

Give it a shot and let me know how it goes...
Look, I don't want to keep repeating myself, so here’s the full breakdown.

BTW, I don't care how many times I say it—you aren't gonna get anything out of this. Or rather, you'll get a tiny bit, but it's basically nothing.

So if I bump it up even by just 1, it's basically gonna stop botting altogether?
What kind of massive headache would that even cause?

Thomas Hernandez21 said:Got a quick question here 🙂

When I crank up the voltage, does the CPU usually die because the voltage is too high, or is it the heat caused by that extra voltage?

The real answer is both.
Basically, more voltage almost always means more heat...🙂
If you push the voltage way too high, the whole rig just dies instantly. But if you set a voltage the chip can technically handle, but that causes the temps to spike past a critical limit—and you don't have an auto-shutdown enabled—your processor is toast. Even if the temp isn't "critical" yet, say it's sitting at 158 °F instead of hitting some extreme ceiling, it’ll likely give up the ghost eventually after running like that for a while.

That's why you gotta keep it under 122-140 °F MAX.🙂

🙂

Look, you aren't even going to be able to bump up the voltage if you don't have some serious cooling under the hood. And even if you do try to push it, you’ve got to go slow—tiny increments only. Every single processor has its own specific limit for what kind of voltage it can actually handle. Those are just the bare minimums, so if you decide to feed it, say, 1V, it might not look like much on paper, but it could still fry your chip. Honestly, every processor is going to kick the bucket if you don't keep it chilled—you'd basically be leaving it to die from a fever or something equally ridiculous.

William Wood5 said:And don't even think about bumping up the voltage if you don't have some serious cooling. Even then, you gotta go tiny bit by tiny bit, because every single CPU has its own specific voltage limits. If you just slap 1V on it out of nowhere, what do you think happens? Every chip is gonna fry instantly if you aren't cooling it with literal ice or some crazy disease or whatever.

Dude, you're talking absolute nonsense... first off, 1V?😲 Most processors nowadays run at around ~1.5V, and here you are talking about increasing it by a full 1V😲. On any decent motherboard, the minimum step is 0.025, not a massive 1V jump.

Then there's that part where you claim every processor will die if you don't cool it with "breath" or some other ridiculous thing.😲😲
My god, the "advice" people give on this site is just pathetic.

Look, you clearly didn't get what I was saying. I know exactly how overclocking works—I know how to push clock speeds and how to ramp up voltage. You don't need to give me a lecture on the basics. You're sitting there talking about "minimal" voltage increases, while I'm telling this guy not to just toss an extra volt in there blindly without thinking. Because, honestly, what does one volt even mean in the grand scheme? You get where I'm coming from, right? Just look at the extremes people go to for cooling—from some guy literally submerging components in mineral oil to the absolute madmen using liquid nitrogen to cool chips that are pushed way past their limits.

William Wood5 said:You can't just crank up the voltage without decent cooling. Even if you do, you have to go extremely slow. Every processor has its own specific voltage tolerance limits. If you push it to 1V, it might not look like much, but it could still kill the chip. Every single one will fail if you don't cool it down properly—you'd basically need to keep it on life support.

Unbelievable. Are you actually serious right now? Where on earth did you even come from?😕

William Wood5 said:Look, you clearly didn't catch my point. I know exactly how overclocking works—how clocks scale and how voltage behaves. You don't need to lecture me. You're talking about minimal voltage increases, but I'm telling this guy that we aren't just bumping it up by a single volt on a whim. You know what I mean? Just look at how people cool processors; from submerging components in oil to using liquid nitrogen for extreme overclocking. It's all out there.

You spend too much time reading Yahoo! and similar garbage. Liquid nitrogen is used for tests that last a few minutes. Please stop talking nonsense. If you actually understood overclocking, you wouldn't be saying this; it's an embarrassment to those of us who actually take this seriously, myself included. 😛

It’s nauseating to read stuff like this. Seriously, people, just because this is Reddit doesn't mean you can peddle such nonsense here.😕

I can see how much you’re struggling when people start pointing out the past—honestly, I already told you to try building a Socket A rig for someone if you think that's going to stay stable indefinitely. Just take a look at this before you go crazy with the overclocking: http://www.markusleonhardt.de/oelbilder.html

Look, I’m not interested in arguing with someone like you, so I’m out. If anyone else needs advice, I'm around, but this is just too much.

If you had actually seen that, you probably would’ve claimed I was being picky too—but honestly, people do some pretty crazy stuff. As for the choice between the 939 and the AM2, I'm leaning toward the AM2. In a year or two, my girlfriend won't have to scrap her entire rig; she'll just be able to swap out half the components, which makes way more sense financially. If you don't want to look at it that way, hey, that's your call.

Honestly, I didn't even bother looking. I already know every CPU cooler on the market, so I don't care what's in that link. Maybe Vappoch?

Besides, this isn't a forum for extreme overclocking. I have no idea why you're posting such nonsense. You have access to XS, so just go join them and try to act smart there. Let's see how long they put up with you.😉

By the way, why replace half the rig or the whole thing? This setup will easily serve its purpose for more than two years. Just check the system requirements; nowadays, even P3 builds get decent upgrades, so there won't be an issue.

I stumbled upon something pretty wild... you guys seriously need to check this out...

http://www.youtube.com/watch?v=KglgLwBE4mY