Do you know why?

Jacolla · 22 Jun 2019 at 03:59

Hi there, this is my first post and i'm not sure if i should do a presentation or something, sooo... here we go.

Well I have an i5 2500k... yehh a good beast, 4 cores & 4 threads, with a little oc to 5gh, man it works pretty good, i've been looking and more or less it can play with some ryzens (in games of course, in multitasks, i just can grab a cigg and watch how he humiliates me...), yeh i know the others don't use oc and that i know, but mine is 8 years old... not bad at all.

Well the point is, i have the oc to 5gh stable and the temps (in "lazy" mode) are around 35-37, in full power total maximum total hits the 83º-85, pretty hot yeh... i cook the breakfast there..., BUT, only in the cores 1 & 2, the 0 & 3 are ALWAYS less than the middle's one's... that's what i'm asking... Do you know why?

Rroff · 22 Jun 2019 at 04:03

Lots of potential reasons - sensors might be calibrated differently, contact between the die and IHS might differ or the contact/thermal material between the IHS and CPU heatsink, or even some differences in the quality of the silicon.

Have you checked which cores are most active during boosting? or is it an all core overclock?

Jacolla · 22 Jun 2019 at 04:15

hmmm, i'll try to replace it and by the way i'll clean this bad girl, thanks man

and they all work at 100% during the work

doyll · 22 Jun 2019 at 12:33

What case you using and what is your case fan setup?

Keep an eye on your CLC. If it's been in use for as long as system has been it's running on borrowed time, and when it goes it's almost always the pump so no cooling until it's replaced.

Jacolla · 22 Jun 2019 at 12:47

Usually i keep an eye on the hole pc, you know, try to keep it clean, and working properly and now the time to a clean deep has come so... by the way i'll have a look at the procesator to check if it was placed good and that, thanks guys, but still with the question haha

doyll · 22 Jun 2019 at 13:47

I wouldn't move CPU. It's working so it has to be seated into socket correctly. Maybe do a new TIM install of cooler. But I'm thinking your higher load temps might be caused to some extent by case airflow not supplying air at similar temp as room .. that the air is being warmed up by heat coming off of components .. and every degree warmer the air becomes translated into almost exactly the same degrees hotter components will be.

You might find the spoilers in link below of interest. It's a basic guide to getting good case airflow:
https://forums.overclockers.co.uk/posts/25142977/

Jacolla · 22 Jun 2019 at 13:53

Well could be, but i think that the hig temp's are because i'm pushing the cpu with 1.6V...

doyll · 22 Jun 2019 at 15:44

Probably correct for CPU as it's under a CLC, but what about air temp into GPU, especially if CLC has fans intaking air from room. Also keep in mind even a new CLC pump has a maximum flow rate between 34L/hr and 68L/hr (about the same as a healthy adult) and it becomes less and less as the years go by .. compared to custom loop pumps being 450-1500L/hour. What this means is heat can build up in waterblock because coolant flowrate isn't fast enough to move it way to radiator. You can check this out by feeling the hoses near pump to see if one is significantly hotter than other.

Yadda · 22 Jun 2019 at 17:47

Jacolla said:
Well could be, but i think that the hig temp's are because i'm pushing the cpu with 1.6V...

1.6v is very high for that CPU, even for a "Doyll approved AIO".

The "middle" two cores heat up more because they each have two neighbouring cores so will naturally get hotter than the two outer cores (which only have one neighouring core).

Jacolla · 22 Jun 2019 at 19:23

Yadda said:
1.6v is very high for that CPU, even for a "Doyll approved AIO".

The "middle" two cores heat up more because they each have two neighbouring cores so will naturally get hotter than the two outer cores (which only have one neighouring core).

haha yehh, it's pretty high... but the cpu still handle it, since a few years ago... this cpu it's massive!

and man, that answer makes a lot of sense, thanks for your answer man

Yadda · 22 Jun 2019 at 19:33

Jacolla said:
haha yehh, it's pretty high... but the cpu still handle it, since a few years ago... this cpu it's massive!

and man, that answer makes a lot of sense, thanks for your answer man

You're welcome, buddy. I hope you've taken precautions - quarter-inch steel plate under your PC should prevent meltdown.

Jacolla · 22 Jun 2019 at 20:37

Yadda said:
You're welcome, buddy. I hope you've taken precautions - quarter-inch steel plate under your PC should prevent meltdown.

haha nahh, i just put the computer under a pan and i can use it to cook and warm myself

Rroff · 22 Jun 2019 at 20:50

Yadda said:
1.6v is very high for that CPU, even for a "Doyll approved AIO".

The "middle" two cores heat up more because they each have two neighbouring cores so will naturally get hotter than the two outer cores (which only have one neighouring core).

Didn't realise Sandy Bridge was laid out like that - I would have assumed they'd be radial layout like many other CPUs. (I probably did know at some point and forgot).

Yadda · 22 Jun 2019 at 21:10

Jacolla said:
haha nahh, i just put the computer under a pan and i can use it to cook and warm myself

We should do a sweepstake as to how much longer it lasts. We could call it "the Jacolla tombola".

Rroff said:
Didn't realise Sandy Bridge was laid out like that - I would have assumed they'd be radial layout like many other CPUs. (I probably did know at some point and forgot).

Yea, when he mentioned "middle ones" I had to look it up too.

Jacolla · 23 Jun 2019 at 02:08

haha, let's see how long it lasts, i believe in her!!

by the way, there is a benchmark i did to... you know, to prove it and that, there is not the voltage... is not that deep, you will have to trust me haha.

https://www.userbenchmark.com/UserRun/17850565

Yadda · 23 Jun 2019 at 02:45

Jacolla said:
haha, let's see how long it lasts, i believe in her!!

by the way, there is a benchmark i did to... you know, to prove it and that, there is not the voltage... is not that deep, you will have to trust me haha.

https://www.userbenchmark.com/UserRun/17850565

You ran a benchmark? I was wondering why my lights flickered.

Rroff · 23 Jun 2019 at 03:20

Some CPUs are surprisingly resilient to voltage - not that I'd push them too far if I was concerned about replacing or recommend anyone did it normally but I ran a Q6600 under water at the max possible voltage table for like 2 years or something of 24x7 use fine then my dad used it (at stock voltage) until recently when I replaced his system.

What tends to kill them in my experience is if you push things like VTT too far.

Jacolla · 23 Jun 2019 at 03:30

Yadda said:
You ran a benchmark? I was wondering why my lights flickered.

hahaaa, jezz, have you notice? i had to use sunglasses cuase was shining like hell... hahahaa

Jacolla · 23 Jun 2019 at 03:32

Rroff said:
Some CPUs are surprisingly resilient to voltage - not that I'd push them too far if I was concerned about replacing or recommend anyone did it normally but I ran a Q6600 under water at the max possible voltage table for like 2 years or something of 24x7 use fine then my dad used it (at stock voltage) until recently when I replaced his system.

What tends to kill them in my experience is if you push things like VTT too far.

man you know, it's pretty old, it's a fighter, this cpu MUST fight until he dies, as a warrior... fighting.

Yadda · 23 Jun 2019 at 03:51

Jacolla said:
hahaaa, jezz, have you notice? i had to use sunglasses cuase was shining like hell... hahahaa

Haha.

A welding mask might be safer.

I can understand the appeal though. Once upon a time I had a P4 1.8GHz that I ran at 2.7 (I think) with a good chunk of unhealthy over-voltage. It eventually died (the only CPU I've ever had die on me) but it was fun and lasted quite a few years.