4670k & 4770k purchased anwhere operating performance thread

Much better - how did you get the GFlops up if you already had SP1?

If you really want to stress the chip, download the newest version of Linpack from Intel - it uses AVX2 and is much harder to pass.
 
v27.9, Blend - custom (~90% RAM), for at least 12 hours

I can pass Intel Linpack (even their latest one) but it will fail in Prime95 after 5 hours.

I will try that some time.

I trust Intel coded avx2 linpack bench mark over prime 95. Prime95 needs avx2 update.
 
Ok guys. The moment has come. Time to shut down and delid.
 
I can pass Intel Linpack (even their latest one) but it will fail in Prime95 after 5 hours.

the batch file for linpack doesn't take that long to execute. if you ran linpack manually setting it to loop a problem size of say, 25000 or 30000, for hours it would probably fail too.
 
SonDa you better hurry! I'm getting antsy.


I got out of The LID.


THEELID.jpg




Into the light.


4770kseesthelight.jpg



All razors. :)


So far everything is going well. Getting ready to mount.

:cool:
 
What did you use on the onboard vrm's? Can't wait to see your results!
 
Much better - how did you get the GFlops up if you already had SP1?

If you really want to stress the chip, download the newest version of Linpack from Intel - it uses AVX2 and is much harder to pass.

SP1 wasn't installed. I did a clean windows install a few weeks ago when I got this cpu and motherboard and forgot to update to SP1. I have the latest version of linpack and have run it before reinstalling SP1. gonna give it a go now. I just found out that this motherboard has a dead pci-e slot so I am replacing it soon. I hope by august 1st microcenter has the gigabyte z87x OC in stock. I really wanted to get that board in the first place.
 
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.



Delidded no IHS direct core mount.
Same water cooling as before.
Same BIOS settings as before.
Coolaboratory Liquid Pro TIM between DIE and Block.


4770KDelidGO.jpg


http://valid.canardpc.com/2865214

Process Gflop performance is up!
With ambient temps taken into consideration temps dropped at least 29C per core!
Also important to note that 100C TDP heat was topping out and the CPU slows dow a little by itself to keep from heating up. So max operation temps were never seen for the before delid test.


Conclusion:

Delidding was a big win for my hot 4770k.
 
Last edited:
Those are incredible results. I wonder if there was a manufacturing defect on your chip. It just sucks that there would have been no real way to go about doing anything about it.

What are your thoughts?
 
Those are incredible results. I wonder if there was a manufacturing defect on your chip. It just sucks that there would have been no real way to go about doing anything about it.

What are your thoughts?

No defect, this is typical of delids. I'm pushing 1.35v and my temps don't go over 70c under any stress test.
 
Those are incredible results. I wonder if there was a manufacturing defect on your chip. It just sucks that there would have been no real way to go about doing anything about it.

What are your thoughts?

No defects. It was kind of the same with my old 3770k. After delidding, my temps dropped by 25c under the same scenario before delid.
 
Those are incredible results. I wonder if there was a manufacturing defect on your chip. It just sucks that there would have been no real way to go about doing anything about it.

What are your thoughts?


Pleased with the results. This is the 3rd Intel chip I have delided. My first was a 3570k and then a 3770k. I have seen good results with each chip that I have delidded.

Delidding has a 100% success rate for me for helping lower operation temps.


It would be great if Intel sold chips without the IHS on them.
 
Pleased with the results. This is the 3rd Intel chip I have delided. My first was a 3570k and then a 3770k. I have seen good results with each chip that I have delidded.

Delidding has a 100% success rate for me for helping lower operation temps.


It would be great if Intel sold chips without the IHS on them.

How did you mount the block to the chip? I'm assuming a special mount is needed? I have an Ek supremacy...what mount do I need to use this block with a delid?
 
How did you mount the block to the chip? I'm assuming a special mount is needed?


I don't think any special mount is needed but I did replace my stock socket back plate with an EK TRUE back plate and I removed the socket cpu retention bracket.

I recommend using CoolLaboratory Liquid Pro TIM since it spreads so thin and then apply gentle hand pressure for mount.

I have done the same mount with 3570k and 3770k.


Light weight copper block DT Sniper block.

DTSNIPERSIDEDirECTDiEmOUNT3570k.jpg


using a EK TRUE 115X back plate for stability.

EKS155TRUEbackplateDTSNIPER.jpg
 
Last edited:
I5 4670K
L314A985 Malay
Cooled by an H110

4300 Mhz @ 1.218 is it for me, Temps are just to high for any more voltage.



The Linpack with AVX 2 is a beast. The Max Temps are with it, I have Prime95 running in the background so you can the difference in temps.
 
The Linpack with AVX 2 is a beast. The Max Temps are with it, I have Prime95 running in the background so you can the difference in temps.



It's the most intense benchmark that I know of that uses AVX2 so it seems appropriate to test haswell with it.
 
While Linpack will produce 100C temps it doesn't crash my computer like Prime 95 will with unstable overclocks.

Anyways I just got another 4670k recently and it seems to be a little bit better than my last one. It doesn't pass the 46 multiplier test, but only uses 1.05v in bios vs 1.15v of the last one. So far it will do 4500 where the last one would only do 4400.
 
While Linpack will produce 100C temps it doesn't crash my computer like Prime 95 will with unstable overclocks.

Anyways I just got another 4670k recently and it seems to be a little bit better than my last one. It doesn't pass the 46 multiplier test, but only uses 1.05v in bios vs 1.15v of the last one. So far it will do 4500 where the last one would only do 4400.


Yes Prime and Intel TAT are more intensive, and they produce slightly higher temps with decent cooling.

Seems the oldies are still the best imo
 
5Ghz? I am only seeing 4999.82:p..

JK, That is awesome..Is is stable @ all? What is the highest you can run stable? 4.7-4.8? If so, what is the Vcore difference needed between your highest stable o/c and 5Ghz?

Still tweaking pushing for 5ghz futuremark benchmark stable. 4.7ghz is the wall for linpack avx2 acceptable temps. I think my mb may have a problem with running chip extra hot.
 
Still tweaking pushing for 5ghz futuremark benchmark stable. 4.7ghz is the wall for linpack avx2 acceptable temps. I think my mb may have a problem with running chip extra hot.

Makes sense...Those VRMs have to be toasty..How is the airflow around the VRM 'sinks? Perhaps watercooling them might very well be worth it when pumping that much current through them, and might very well net you that stability..
 
Makes sense...Those VRMs have to be toasty..How is the airflow around the VRM 'sinks? Perhaps watercooling them might very well be worth it when pumping that much current through them, and might very well net you that stability..



I got the EK VRM and chipset heat sinks for my MB last week. Haven't had time to mix them in my loop yet. I hope they help.
 
Last edited:
I just wanted to amend my post above.

4300 Mhz @ 1.218 was NOT prime stable. It took 1.232 to achieve that.

I have 4400 Mhz stable @ 1.266 with the bios that shipped with my Z87 Sabertooth.

After reading this thread about the bios update, http://hardforum.com/showthread.php?t=1770773&page=3 , I updated and am now testing 4500 Mhz @ 1.266.

I will report back later.
 
5Ghz? I am only seeing 4999.82:p..

JK, That is awesome..Is is stable @ all? What is the highest you can run stable? 4.7-4.8? If so, what is the Vcore difference needed between your highest stable o/c and 5Ghz?

I'm at a manual 1.41 to boot at 5GHz, but my DDR is at 2133. It'll crash pretty quickly with a small workload, looking at more like 1.45 to 1.47 to start running stuff, but temperatures are astronomical.

4.8 is stable with a 0.225v offset, at least as far as gaming goes. Looking at 1.33v while running games in that case, and heat isn't an issue. Can't run stress tests at all due to heat (sits there throttling at 100C). Lately I've been running at 4.7 with a 0.160v offset, which yields about 1.25v during games. I can also run some stress tests, although it'll jump to 1.374v for AVX loads. A manual 1.325v also works for 4.7, allowing for more stress tests.
 
Perhaps watercooling them might very well be worth it when pumping that much current through them, and might very well net you that stability..


Done! :cool:


Z87X-OCblocksonline.jpg


Water cooling will help with the cooling but I'm not sure if that will help out with the over clocking. I will soon find out.
 
So close to being where I would like to be with my Z87X-OC.

Ambient Temp water cooling.

Haven't been on [H] in almost a year, great to see SonDa5 is going hard on Haswell/Z87 with some really nice gear and 5GHz results :) Keep it up dude.
 
Very nice results. My 4670K is prime95 stable for hours on end at 1.226v/4.5ghz and I wonder if that makes it a good candidate for delidding?
 
Very nice results. My 4670K is prime95 stable for hours on end at 1.226v/4.5ghz and I wonder if that makes it a good candidate for delidding?

Delidding greatly helps with cooling. If your temps seem extremely hot for its speed and voltage yet it still clocks fast with low voltage you have a good chip to delid.
 
Back
Top