4670k & 4770k purchased anwhere operating performance thread

4770k here, Malay batch# L312B323. I am getting a stable 4.5GHz @ 1.25v core (adaptive), 1.18v ring (adaptive), 45x multi on core, 42x multi on ring.

Cannot seem to get her stable at 4.6 and unwilling to go past 1.30v as AIDA64 temps spike to 100c and CPU begins to throttle.

I was a little disappointed at first but the more reading I do it seems I actually have a decent chip. And yes, at 1080p 120Hz I certainly AM noticing a difference in gaming framerates with dual 7970's compared to my old 4.13GHz i7 920.

Great chip and platform, I am glad I made the upgrade!
 
NukeDukem said:
4770k here, Malay batch# L312B323. I am getting a stable 4.5GHz @ 1.25v core (adaptive), 1.18v ring (adaptive), 45x multi on core, 42x multi on ring.

Cannot seem to get her stable at 4.6 and unwilling to go past 1.30v as AIDA64 temps spike to 100c and CPU begins to throttle.

I was a little disappointed at first but the more reading I do it seems I actually have a decent chip. And yes, at 1080p 120Hz I certainly AM noticing a difference in gaming framerates with dual 7970's compared to my old 4.13GHz i7 920.

Great chip and platform, I am glad I made the upgrade!
Click to expand...

From what I have read AIDA64 doesn't stress the cpu very well.


If you want max stress load and temps go with AVX2 linpack benchmark.
 
Here are my results for my maximum performance over clock on my 4770k before I delid it.

Powered by Intel's latest AVX2 linpack code.

Look at the Gflop performance and the heat.

4770k with HT off at 4.4GHZ.
Custom water loop, not delidded yet
Ambient temps 24C

AVX2HToff.jpg


http://valid.canardpc.com/2859272


This chip is worthy and ready for delidding.
 
Last edited:
SonDa5 said:
This chip is worthy and ready for delidding.
Click to expand...

Wow, only 4.4Ghz without HT and you are still hitting 100C on several cores under full water?:eek::eek::eek:...

Man, Hasfail is really loving up to its name..I really, really, hope Intel does a respin and can pull out a Q6600 G0 stepping miracle on the 4770Ks..
 
ccityinstaller said:
Wow, only 4.4Ghz without HT and you are still hitting 100C on several cores under full water?:eek::eek::eek:...

Man, Hasfail is really loving up to its name..I really, really, hope Intel does a respin and can pull out a Q6600 G0 stepping miracle on the 4770Ks..
Click to expand...


The temps are pathetic and delidding is in order. I hope a new stepping miracle for 4770k happens as well.
 
ccityinstaller said:
Wow, only 4.4Ghz without HT and you are still hitting 100C on several cores under full water?:eek::eek::eek:...

Man, Hasfail is really loving up to its name..I really, really, hope Intel does a respin and can pull out a Q6600 G0 stepping miracle on the 4770Ks..
Click to expand...

Gotta agree, my chip is awful but those temps seem insanely high even for what I've been seeing. I've only seen the chip throttle at about 1.3v, and that's with hyper threading on (using a Corsair H100i).

Although that new linpack is a bitch. Seriously, I'm not even stable at 4.2Ghz and 1.25v core using it. Screen goes black and computer restarts within a few minutes. I thought my chip was bad, but wow. Not sure I'm comfortable pushing 1.3v for just 4.2Ghz... :eek:

I'm also curious, how are you running linpack in the screenshot above? Are you using LinX, or something else?
 
Anyone here have a 4670k paired up with a Hyper 212 Evo? Could you PM me. I want to compare temps!

Thanks,
 
SonDa5 said:
Here are my results for my maximum performance over clock on my 4770k before I delid it.

Powered by Intel's latest AVX2 linpack code.

Look at the Gflop performance and the heat.

4770k with HT off at 4.4GHZ.
Custom water loop, not delidded yet
Ambient temps 24C

This chip is worthy and ready for delidding.
Click to expand...

Wow, now that is hot :eek:.

It's going to be interesting to see on how much of a drop you'll get.
 
Eriksrocks said:
Gotta agree, my chip is awful but those temps seem insanely high even for what I've been seeing. I've only seen the chip throttle at about 1.3v, and that's with hyper threading on (using a Corsair H100i).


I'm also curious, how are you running linpack in the screenshot above? Are you using LinX, or something else?
Click to expand...


Keep in mind that my 4770k is being stressed to its performance and thermal limits with this linpack.

Look at the insane Gflop output.

Also keep in mind that my memory is at 2400mhz CL9.

My results truly capture hardcore CPU stress testing that only AVX2 powered cpus are capable of.


I'm running the runme_xeon64 linux file that is in the linpack bench folder. Has a test settings ready to go. You just click it and go and monitors temps and load and wait for it to run its course. Take around 10 minutes to run but you will not see anything. The results are kept in the linpack benchmark folder titled win_xeon64. Open up the text and you will see the results and that is how I took the screen shot with CPU-Z open and REal temp.
 
SonDa5 said:
Here are my results for my maximum performance over clock on my 4770k before I delid it.

Powered by Intel's latest AVX2 linpack code.

Look at the Gflop performance and the heat.

4770k with HT off at 4.4GHZ.
Custom water loop, not delidded yet
Ambient temps 24C

AVX2HToff.jpg


http://valid.canardpc.com/2859272


This chip is worthy and ready for delidding.
Click to expand...


Your water cooler is doing a very poor job almost air cooling performance, my modest Corsair Hydro 110i maxes 82C running prime indefinitely & I am forced to run 4.4Ghz 1.330 vcore, a fair bit higher than you. Something not right there!
 
Battleneter2 said:
Your water cooler is doing a very poor job almost air cooling performance, my modest Corsair Hydro 110i maxes 82C (22 Ambient) running prime indefinitely & I am forced to run the 4770K @4.4Ghz 1.330 vcore a lot higher than you. Something not right there!
Click to expand...



Run the same benchmark I ran with your 4770k at 4.4GHZ with memory at 2400mhz CL 9 and post a screen shot if you pass the test.

This linpack benchmark is bad ass at heating up cores. Also I disable all the power saving features in the BIOS so that my cpu doesn't take a break when working.
 
Battleneter2 said:
Your water cooler is doing a very poor job almost air cooling performance, my modest Corsair Hydro 110i maxes 82C (22 Ambient) running prime indefinitely & I am forces to run the 4770K @4.4Ghz 1.330 vcore a lot higher than you. Something not right there!
Click to expand...

Powered by Intel's latest AVX2 linpack code.
Click to expand...

As he and others have mentioned, this will torture the cores a lot more.
 
sorry for the multiple posts guys was trying to correct a error and the page wasn't loading :p



SonDa5 said:
Run the same benchmark I ran with your 4770k at 4.4GHZ with memory at 2400mhz CL 9 and post a screen shot if you pass the test.

This linpack benchmark is bad ass at heating up cores. Also I disable all the power saving features in the BIOS so that my cpu doesn't take a break when working.
Click to expand...


All my power saving features are off, never use them. Intel TAT gets to the mid 80's which has a rep as one of the worst if not worst CPU heat stress test. As I say my Vcore is night and day higher than yours (has to be), your ram speed will have a minor effect.

I will run Linpack later in the day and let you know, but I strongly suspect something is off with your cooling.
 
Battleneter2 said:
A small difference, I bet if you clock it down to 1600 you wont see much of a core temp drop. Do it!
Click to expand...



I can do that but I may wait till after I delid. I think if I lower my memory speed and memory voltage I should also be able to lower vcore on cpu. Im sure temps will come down.
 
Battleneter2 said:
Your water cooler is doing a very poor job almost air cooling performance, my modest Corsair Hydro 110i maxes 82C running prime indefinitely & I am forced to run 4.4Ghz 1.330 vcore, a fair bit higher than you. Something not right there!
Click to expand...



Here is something else to think about.


The problem with the IHS and TIM that Intel is using is related to the offset of the IHS and proper application of TIM from the factory.

I am thinking that part of the reason my temps are so high for the low voltage that I am using to push 4.4GHZ is because Intel screwed up when they set the IHS on the die. I'm thinking there is huge space and air bubbles and bad contact in between the die and the IHS.

Once I delid and get a direct bare die to water block mount the fix should be noticeable in a very nice drop in temps. I have seen people that delid drop anywhere from 15C to a little over 30C once delidding and getting a good mount with a decent TIM.
 
SonDa5 said:
Run the same benchmark I ran with your 4770k at 4.4GHZ with memory at 2400mhz CL 9 and post a screen shot if you pass the test.

This linpack benchmark is bad ass at heating up cores. Also I disable all the power saving features in the BIOS so that my cpu doesn't take a break when working.
Click to expand...

Well I ran Linpack, can't say I am a fan unless I am missing something (entirely possible). Intel TAT is even worse as it uses so much processor the mouse jitters.




Now I didn't compare these very long (saw no reason to go longer) and its first thing in the morning here so ambient temps are like 15C for reference. Just adding looks like one can customise the batch files in Linpack but why bother when Prime or TAT does the job with one click.

Your Intel theory is possible but I am a logical person and I would suggest its more likely your block is not sitting flat or you have some sort of restricted flow issue.
 
Last edited:
Battleneter2 said:
Your Intel theory is possible but I am a logical person and I would suggest its more likely your block is not sitting flat or you have some sort of restricted flow issue.
Click to expand...



My custum loop is in good operating order. What was your Gflop performace? Can you post screen shots of the result text file?
 
SonDa5 said:
My custum loop is in good operating order. What was your Gflop performace? Can you post screen shots of the result text file?
Click to expand...

You are over thinking it, your temps are NOT consistent with most other people on half decent liquid cooling for 4.4Ghz 1.2 volts, they are god awful. Your loop has issues or is not very good take your pick, it 95% likely one or the other,

My temps are consistent with most others on mainstream liquid coolers. I can lead you to water but no one can make you drink.

Clock your ram down to 1600 (for you own piece of mind, I know its not the issue), run Intel TAT/prime 95 I would bet my house you run higher temps again, confirm your cooling is sucking.

Complete the above steps and we will talk again
 
Battleneter2 said:
You are over thinking it, your temps are NOT consistent with most other people on half decent liquid cooling for 4.4Ghz 1.2 volts, they are god awful. Your loop has issues or is not very good take your pick, it 95% likely one or the other,

My temps are consistent with most others on mainstream liquid coolers. I can lead you to water but no one can make you drink.

Clock your ram down to 1600 (for you own piece of mind, I know its not the issue), run Intel TAT/prime 95 I would bet my house you run higher temps again, confirm your cooling is sucking.

Complete the above steps and we will talk again
Click to expand...


I have peace of mind. I have done tests and my chip is hot but its decent with voltage. My problem isnt the loop. The problem is the chip itself and its more than likely related to poor ihs mount from factory. Delid and kick ass mount should cool it way down.
 
SonDa5 said:
http://valid.canardpc.com/2862532
[email protected], 1600 memory. HT off.
Didn't pass all test



Temps a little cooler.

With the 2400mhz the system finished in alot less time and gflop performance was higher. Temps were only slightly less.

Looks like this chip suffers from a bad mount from the factory.
Click to expand...

At least you eliminated the ram overclock as the primary cause.

Why are you hell bent on refusing to consider your cooling is the issue?, just reseat your block for a start.
 
I'm starting to get a hang of how this CPU overclocks. The same goes with my Samsung RAM. :)




EDIT: Battleneter2, when you mention your temps with Prime95, are you using version 27.9?
 
Battleneter2 said:
At least you eliminated the ram overclock as the primary cause.

Why are you hell bent on refusing to consider your cooling is the issue?, just reseat your block for a start.
Click to expand...


This is my 2nd mount. Not sure what kind of TIM I am using since it came in an unmarked container from one of the water blocks I have accumulated over the years. It looks like it is Shin-Etsu. It looks identical to this but unmarked. http://www.diygadget.com/accessories-1/shin-etsu-x23-7762-thermal-compound-1-gram.html

I have Cool Laboratory Liquid Metal Pro but I don't want to use it on the IHS because it will probably damage the markings. When I delid I am going to use Cool Laboratory Liquid Pro with a direct mount without the IHS.

I have been using this block for awhile. I had it working great with a delidded 3770k. I haven't changed up my loop flow order. My GPU is in the same loop as before and the GPU cools great.

It's just the CPU that is running hot.

I really think the problem is that the IHS is not mounted well from the factory. I don't think the problem is related to my cooling.


I am going to do a 3rd mount. This time I will use a different TIM. I found some XSPC K2 laying around which is supposed to be almost as good as MX4. http://www.vortez.net/articles_pages/xspc_raystorm_750_rs360_watercooling_kit_review,11.html

After 2 mounts I am convinced that my chip runs hot. A 3rd mount with different TIM will nail the final verdict down. After this 3rd test I am done doing pre-delid testing. I want to move forward and fix my hot 4770k by delidding it and applying a kick ass direct to die mount.
 
Last edited:
3rd and final pre-delid test with linpack with AVX2.
4770k 4.4GHZ 1.2v, Memory 1600mhz http://valid.canardpc.com/2862987
Ambient Temp 24C
TIM XSPC K2
3rd mount

3rdtestpredelidXSPCK2TIM.jpg


Did not pass problem size 35000. With memory slower than 2400mhz the speed of the benchmark is much slower. Gflop performance is slightly less.

Temps are still hot. Proper delid is in order.
 
Battleneter2 said:
Well I ran Linpack, can't say I am a fan unless I am missing something (entirely possible). Intel TAT is even worse as it uses so much processor the mouse jitters.

Prime95vsLinpack.png



Now I didn't compare these very long (saw no reason to go longer) .
Click to expand...


The benchmark takes a few moments to begin to heat up the CPU. 15C ambient temp is much cooler than mine. (Mine are 24C)


For best results disable HT in the BIOS and your cpu can focus on the benchmark and only use 4 cores. This benchmark works best with 1 thread per core.
 
SonDa5 said:
The benchmark takes a few moments to begin to heat up the CPU.
Click to expand...

i've found the quickest way to get the heat up is to not use the batch file but instead run the .exe directly in interactive mode.

double click .exe, skip the first prompt (press enter), problem size -> 25000, leading dimension -> 25000, number of trials -> 1, data alignment -> 4

one trial is enough to heat the cpu up but if you want to run it longer you can increase the number of trials.

with these settings my 4670k starts throttling in about 40 seconds.
 
Batch = L315B352
Made in Malaysia
Asus z87 pro
2x8GB Corsair vengeance CL9
Thermalright ultra E 120

1.2vcore + 4.4 = No go.
1.25vcore + 4.5 = No go.


vcore = 1.175 at 4.3
Cache = 4000
HT enabled

P5DJmBz.png
 
Batch # L311B288
Malaysia
Asrock Z87 Extreme 6
8GB 1600 mhz ram (now have a 2400mhz set)
H100i

4.5ghz @1.199v stable (cache ratio 34/auto)

lvrD5bf.png


I can see myself changing the motherboard and maybe getting a 4770k in the near future. that's why I haven't tried to de-lid yet.
 
PureBlackFire said:
Batch # L311B288
Malaysia
Asrock Z87 Extreme 6
8GB 1600 mhz ram (now have a 2400mhz set)
H100i

4.5ghz @1.199v stable (cache ratio 34/auto)

lvrD5bf.png


I can see myself changing the motherboard and maybe getting a 4770k in the near future. that's why I haven't tried to de-lid yet.
Click to expand...

You need to make sure you have Win 7 SP1 installed - it doesn't look like IBT is using AVX extensions in your screenshot, that's why the GFlops are so low. Should be 100+.
 
thanks for the info Forceman. i always wondered why it was so low. is there a way i can check if it's enabled? I'm certain I have SP1.
 
You must log in or register to reply here.
Back
Top