Too hot on water. Why?

[MxPhenom 216]

I just dont think that 240 radiator of the h220 can handle the heat off of a 6 core chip at 4.5ghz and 1.4v. Id also like to know the temps at stock clocks and voltage.

TIM doesnt seem to be spreading well eitjer. That IC Diamond stuff is pretty lack luster TIM.

 

[reppel]

I just dont think that 240 radiator of the h220 can handle the heat off of a 6 core chip at 4.5ghz and 1.4v. Id also like to know the temps at stock clocks and voltage.

TIM doesnt seem to be spreading well eitjer. That IC Diamond stuff is pretty lack luster TIM.

Read my previous post

 

[reppel]

I've now removed the block again, used the recommended thermal paste that came with swiftech, shaken the block and rad in case of air bubbles, repositioned the block with the tubing upwards, repositioned the radiator with the reservoir opening upwards, same problem overall, same temperatures, although strangely less stable. I was running 4.6ghz at 1.3 which now got unstable.
I don't know whether the problem is with the swiftech thing or the cpu.

Running at STOCK speed, case open, fans and pump at max speed, prime95 during 5 minutes:

EDIT: Ambient 22 degrees

 

[Kaynar]

Hm ok, by the way I checked my screenshots I have 54c hottest core under max load at stock speed (3.9ghz) after about 10mins of stress test (on previous post I said 67c, but I was wrong, 67c is the CPU temp of my laptop under load lol). Bear in mind my desktop has a custom loop though, as I described earlier. Unfortunately I am not home these days so I can't tell you my voltage or any other info, but your stock temps seem logical for an AIO therefore I think your H220 has no issues and you just have a terrible 4930k for oc. I mean... more than 1.2v at stock seems kind of high...

(http://i931.photobucket.com/albums/...259ethigrave16-12-2013-172815_zps686970a0.jpg)

 

[cadaveca]

I think your H220 has no issues and you just have a terrible 4930k for oc. I mean... more than 1.2v at stock seems kind of high...

Based on my IVB-E chips, I'd agree. Quality of these chips can be pretty varied.

At the same time, my "good" chip gets the same temps on H100, or H80i, and only slightly warmer under air-cooling. Some chips just run hot...and others do not.

 

[reppel]

But that can means that if I had a stock cooler that it would run in the 90s. And isn't the stock voltage the same for everyone?

 

[Vario]

The later ivy bridges all seem to have really bad overclockability, my 2013 purchase i5 and two i7's all were horrible. This was microcenter binning.

 

[MxPhenom 216]

But that can means that if I had a stock cooler that it would run in the 90s. And isn't the stock voltage the same for everyone?

No, not at all. That is how people usually determine if they have a good chip or not. Lower the stock voltage typically means you have a better chip. And what a general good stock clock for those chips, I have no idea.

 

[reppel]

I was running at 4.6 stable at 1.29v, stable. I don't know if that is good or not, but my problem is mostly the temps, not the voltages, i think. In that case, the problem was it getting in the 90s.
When I ran stock at 1.2v, it was because it was on auto, i didn't try to run it at lower voltages, but i suppose it would go lower.

 

[Vario]

No, not at all. That is how people usually determine if they have a good chip or not. Lower the stock voltage typically means you have a better chip. And what a general good stock clock for those chips, I have no idea.

I think a rule of thumb is gold chips are 1.05v or below stock voltage frequently. Good overclockers are like 1.15-1.10vish. I think my vid for 4.5 is around 1.18ish for example, and stock runs about 1.15v. I can get to 48x multi easily, haven't tried over 1.4v though.

I think theres two things: stock vid and the distance from the lid to the die (die contact to heatsink).

This is on a 3770k but the Ivy-E can't be that much different.

 

[reppel]

I think a rule of thumb is gold chips are 1.05v or below stock voltage frequently. Good overclockers are like 1.15-1.10vish.

But when you say gold chips at 1.05v, do you mean when you leave it on auto, or when you lower the voltage until you find a minimum stable value?

 

[Vario]

But when you say gold chips at 1.05v, do you mean when you leave it on auto, or when you lower the voltage until you find a minimum stable value?

If you were to boot the machine with all default settings, your stock voltage should be really low if you have a good chip. Bad chips require about 1.25v for the stock speeds. These usually max out at 43-45x.

 

[reppel]

If you were to boot the machine with all default settings, your stock voltage should be really low if you have a good chip. Bad chips require about 1.25v for the stock speeds. These usually max out at 43-45x.

Are you 100% on that? I was under the impression that all 4930k out of the box on stock settings, all have the same voltages, if nothing changed. I mean, this may be a dumb question, but how does the computer know which voltage to give at stock to make it stable?

And take a look here: http://extremespec.net/processor-intel-core-i7-4930k-review-testing-performance/
The first two CPU-Z graphs. The second one, which is at load, has the same voltage as mine, like all out of the box should have, right?
The first graph has only 1.08V, but it's not running at boost speed.

 

[MxPhenom 216]

Are you 100% on that? I was under the impression that all 4930k out of the box on stock settings, all have the same voltages, if nothing changed. I mean, this may be a dumb question, but how does the computer know which voltage to give at stock to make it stable?

And take a look here: http://extremespec.net/processor-intel-core-i7-4930k-review-testing-performance/
The first two CPU-Z graphs. The second one, which is at load, has the same voltage as mine, like all out of the box should have, right?
The first graph has only 1.08V, but it's not running at boost speed.

@cadaveca can tell you more on this. He's the most knowledgeable when it comes to overclocking on this site as far as I know. Chips are all of different quality, one chip is not the same as the other, so they will require more voltage for certain clocks than the other.

Also, I think its within the chip itself, how much the motherboard give for stock voltage, but Dave can probably give more legitimate information on that.

 

[cadaveca]

Are you 100% on that? I was under the impression that all 4930k out of the box on stock settings, all have the same voltages, if nothing changed. I mean, this may be a dumb question, but how does the computer know which voltage to give at stock to make it stable?

All chips are different. They are "programmed" for voltage by Intel when manufactured. You can go to the Intel site and download the spec sheet which lists the possible voltages for all parts of the chip, even, as IMC and CPU can all use different voltages.

Motherboards, on the other hand, tend to set voltages to similar levels regardless of chip, but more on the high side, than the low side. The best way to find your CPU's default voltage is to set everything manually using "offset" settings, and use what is then set to try to figure out what your actual stock voltages are.

 

[theFOoL]

as another user suggested please use the stock cooler and tell us the temps. be sure to run stock or set to auto (since you said the temps were the same when you down clocked back a bit on Voltage and Freq.) if the temps are the same it's the cpu

on the other hand... i have known for Motherboards to have faulty Temp. sensors but this the last thing you want to be the problem

 

[reppel]

All chips are different. They are "programmed" for voltage by Intel when manufactured. You can go to the Intel site and download the spec sheet which lists the possible voltages for all parts of the chip, even, as IMC and CPU can all use different voltages.

Motherboards, on the other hand, tend to set voltages to similar levels regardless of chip, but more on the high side, than the low side. The best way to find your CPU's default voltage is to set everything manually using "offset" settings, and use what is then set to try to figure out what your actual stock voltages are.

Couldn't find that information on spec sheets, but i take your word on that.
I did the offset thing, it seems the same, 1.232. Actually I think it's 1.2 but it's reading more, as every value i set, it shows a tiny bit more.
But regardless of that, let's say I overclock 10 different 4930k's all at 4ghz and 1,4V, which I'm assuming all of them can. Wouldn't they all produce similar temperatures, or do you mean to say that, because they have all different performances, I should expect up to 20-30 degrees difference between them?

as another user suggested please use the stock cooler and tell us the temps. be sure to run stock or set to auto (since you said the temps were the same when you down clocked back a bit on Voltage and Freq.) if the temps are the same it's the cpu

on the other hand... i have known for Motherboards to have faulty Temp. sensors but this the last thing you want to be the problem

I can't unfortunately. I didn't get any cooler on the cpu box, didn't think i would need one either.

I really don't mind much if the cpu doesn't overclock much. I did 4.6 and was pretty happy about that. My issue is that it seems that everyone out there is getting colder temperatures on air, and I was wondering if this was normal or if there was something wrong in my system, or if i did something wrong.

 

[cadaveca]

Wouldn't they all produce similar temperatures, or do you mean to say that, because they have all different performances, I should expect up to 20-30 degrees difference between them?

They are all different. The 4830..quadcore with HT. 4930, hexacore with HT and 12 MB of cache. 4960X, hexacore with 15 MB of cache. Within each model, the silicon can be quite varied, with some chips only capable of 4.3 GHz or so, while others can reach 5 GHz on far less voltage than stock of those poor chips( ie, 1.25V for stock, hit 4.3 GHz max, whil2 another needs 1.25 for 5 GHz, and has much lower stock voltage.)


I have multiple 4960X chips, and confirm that this is definitely the case. Overclock, power consumed, temperatures, and everything are quite different for each individual chip.

 

[reppel]

They are all different. The 4830..quadcore with HT. 4930, hexacore with HT and 12 MB of cache. 4960X, hexacore with 15 MB of cache. Within each model, the silicon can be quite varied, with some chips only capable of 4.3 GHz or so, while others can reach 5 GHz on far less voltage than stock of those poor chips( ie, 1.25V for stock, hit 4.3 GHz max, whil2 another needs 1.25 for 5 GHz, and has much lower stock voltage.)


I have multiple 4960X chips, and confirm that this is definitely the case. Overclock, power consumed, temperatures, and everything are quite different for each individual chip.

Thx, and I just don't want to misunderstand, just to confirm I understood.
The first part you said, they all achieve different results, I understand that and agree.

But what I was asking, and just to make sure I understand, if I put all those ten chips running at EXACTLY the same conditions on the bios. Let's say 4.0000GHz and 1.40000v, all set manually at these values. I'm then assuming they all produce very similar, if not equal, benchmark speed results. But temperature-wise, in these conditions, they all produce different temperatures, ranging 20-30 degrees between them?
(sorry for insisting, just making sure i understand )

EDIT: And if that's the case, and considering that I'm running at around 75degs on water on stock (which are the settings Intel guaranties it can run at). Can't I RMA the CPU because technicaly, if I used the stock cooler, it would run at 90+ and it would throttle, which is not suppose to happen, maybe?

 

[cadaveca]

Nope. I mean ,you could RMA, sure, but these chips don't come with coolers for precisely that reason. I mean ,you can run "stock", I am sure, under the Intel AIO watercooler just fine, maybe 4.2 GHz. RMAs aren't really all that complicated, really.

If you set the clocks the same like that, with that high voltage, yeah numbers will all be the same, provided some of those chips aren't throttling. At the same time, of course temperatures are different, as that's what's silicon as all about, really, no matter what CPU design you can think of...or memory...or anything else that uses silicon as a transistor.

At the same time, with that 1.4V...which is actually pretty excessive for IVB-E in my books, many CPUs are going to be able to OC differently. 1.25V to 1.3V is fine, I am fairly confident, but at the same time, it's not like I really have any real idea as to what's safe or not, other than going by Intel's whitepapers that describe these things.

If you aren't happy with your chip for OC, but want to have a clear conscience about RMA a "working" chip, then buy the Intel Tuning Plan for the $25 or so it costs, wait the 30 days for it to activate, and RMA that way.

 

[reppel]

Thx, that pretty much answers all my questions.
About the intel tuning plan. I read on their conditions that it has to be a boxed cpu and not a tray cpu.
How do I see that?
I mean, the cpu came in a box, but without cooler.

EDIT: something like this one:

EDIT2: Oh, one more thing
You think it works in Sweden?

 

[cadaveca]

Yeah, that's a boxed CPU. Tray CPUs are what companies like Dell and such get, and they have other options for Intel Tuning Plan for the OC'd chips they sell in some systems.

Here's the website, it's probably best to read all the infos there rather than take my word for it.

http://click.intel.com/tuningplan/

 

[reppel]

Thx Dave! You're the best. I'm definitely doing that

 

[radrok]

If it helps I've been running 4,8GHz at 1.42v for a little over a year with no issues but I don't know if Ivy-E is more delicate than SB-E.

This is literally 24/7 because I can't remember how many nights I left this PC rendering projects.

Benched 5,3 GHz at 1.55v several times too!

 

[reppel]

If it helps I've been running 4,8GHz at 1.42v for a little over a year with no issues but I don't know if Ivy-E is more delicate than SB-E.

This is literally 24/7 because I can't remember how many nights I left this PC rendering projects.

Benched 5,3 GHz at 1.55v several times too!

How many radiators you running?
And what max temps you getting?

EDIT: just tested mine at stock with low pump and fans speed, temps went up to 84.