4790k's are old

[lavaheadache]

mine was minted week 31 of 2013. I haven't even installed it yet but just thought I'd share the info.

I'll come back and share my thoughts.

I have had 3 different 4770k's that all seemed to have sweet spots at about 4.3-4.4 before I had to push up the volts more than I'd like.

Hoping that this little devil will finally help me like Haswell.

 

[Techhog]

That's the best you could go on the 4770Ks? My 4670K is stable at 4.2 from just raising the multiplier...

Anyway, this isn't the first time I've heard about a 4790K from 2013 here, but this one seems a bit extreme.

 

[lavaheadache]

the 4770's could go higher. I was just saying that it took a good bump in volts to get to like 4.5. Most would need 1.3-1.35v.

Anyhow, so far my 4790 is doing 4.4 @ 1.168 and load temps 64c peak on the hottest with a Silver Arrow SB-e at auto fan settings

Big improvement across the board so far.

edit********

4.8 @ 1.31 so far with load temps about 80c on the hottest

 

[Homeles]

Here's IDC's comment on the matter:

It's not the package assembly date, it is the lot-start date.

That is, the date at which the wafer was officially laser-engraved with its lot number when the lot was started in the fab.

This is industry standard protocol BTW, everyone does it the same way because this is how QRA tracks the silicon in the field and associates it with its processing history in the fab. (for tracking down root-cause problems and so forth)

A lot start date of Sept 2013 would put the wafer ex-fab around end of Nov to early Dec, after which it goes through test, bin, and then shipped to a packaging plant by boat, packaged and then shipped by boat yet again to a warehouse where it sat until it was trucked to a retailer destination.

It takes time to do this at a global level.

 

[Galatian]

the 4770's could go higher. I was just saying that it took a good bump in volts to get to like 4.5. Most would need 1.3-1.35v.

Anyhow, so far my 4790 is doing 4.4 @ 1.168 and load temps 64c peak on the hottest with a Silver Arrow SB-e at auto fan settings

Big improvement across the board so far.

edit********

4.8 @ 1.31 so far with load temps about 80c on the hottest

What are you using to stress your CPU. I'm at 4,7 GHz but reach 96 degrees C in Prime95...this is with a custom waterloop.

Any recommendations on overclocking. Right now I left everything on Auto and just changed core voltage. I also reduced the ring multiplier to 34. Seems easier to get it stable that way. Maybe I'm just doing something wrong?

 

[rtsurfer]

You should have waited.

L4 is the newest batch & people have reported it to be a better clocker generally.

I also read about a guy who just told the employee at MicroCentre that he would buy the Processor Insurance if he could get him an L4 chip.
The employee dug through the processors he had in the back & got the guy an L4 chip.

Although your OC looks good I am wondering what you are using to test it ?

 

[toyota]

That's the best you could go on the 4770Ks? My 4670K is stable at 4.2 from just raising the multiplier...

Anyway, this isn't the first time I've heard about a 4790K from 2013 here, but this one seems a bit extreme.

so you raised the speed and left the voltage on auto? if thats what you did then of course it was stable as it will feed it more than enough voltage like that. heck I can leave everything on auto and set the speed to at least 4.6 but thats the highest I tried because it goes well over 1.30.

 

[Techhog]

so you raised the speed and left the voltage on auto? if thats what you did then of course it was stable as it will feed it more than enough voltage like that. heck I can leave everything on auto and set the speed to at least 4.6 but thats the highest I tried because it goes well over 1.30.

Yeah, but even then it only went to 1.217V in CPU-Z. (Though, that's a bit much for my cooler and I should try lowering it a bit.)

 

[lavaheadache]

Been testing with prime 95 so far because I'm old like that. After a bit on prime 95 I played bf4 for about 2 hours. Seems rock solid at 1.31 with good temps. I just may stop here and and finalize stability. I didn't even push further

 

[cytg111]

What are you using to stress your CPU. I'm at 4,7 GHz but reach 96 degrees C in Prime95...this is with a custom waterloop.

Any recommendations on overclocking. Right now I left everything on Auto and just changed core voltage. I also reduced the ring multiplier to 34. Seems easier to get it stable that way. Maybe I'm just doing something wrong?

I hope that is the AVX version atleast ?

 

[Galatian]

I hope that is the AVX version atleast ?

Yes version 28.5 I believe

 

[BonzaiDuck]

Has anyone been able to test and quantify the thermal improvements on these Devil's Canyon chips like the 4790K? Supposedly they had addressed the TIM shortcomings with a new "polymer" TIM substance.

 

[SpacemanSpiff46]

Aw man, mine is arriving today. I'll let you know what my chip says and is capable of.

 

[Homeles]

Has anyone been able to test and quantify the thermal improvements on these Devil's Canyon chips like the 4790K? Supposedly they had addressed the TIM shortcomings with a new "polymer" TIM substance.

There's been a handful of reviews. General consensus is that temps are down (or at least flat, at worst), even at stock, despite a higher TDP and ~15% higher stock clock and turbo.

I guarantee that the primary difference with the TIM is a higher thermal conductivity. Wish Intel would release the physical properties of it, who's the supplier (probably Dow Corning again), and the "model number."

Actually... does look like Dow Corning made some sort of improved silicone polymer TIM last year:
http://www.dowcorning.com/content/news/ECTC_2013_Present_Paper_Launch_Compounds.aspx

 

[rtsurfer]

From what I've read the new TIM is 5-10 C better than the old one at similar Clocks & Voltages.

Marginal improvement.
4.8 - 5.0 Ghz still requires de-lid.

 

[Homeles]

From what I've read the new TIM is 5-10 C better than the old one at similar Clocks & Voltages.

Marginal improvement.
4.8 - 5.0 Ghz still requires de-lid.

Yeah, it's just TIM. But that's a pretty big improvement for such a minor change.

 

[campbbri]

Today I picked up a batch L329 from Microcenter, so even older than yours.

I'm one of the few not planning on overclocking so batch numbers don't affect me but it is interesting how large the range of dates is on these chips at launch. With the tight window and the fact they resorted to a "paper" launch, I thought they would all come from a few recent batches.

 

[WhoBeDaPlaya]

Come to think of it, where the heck is AT's review of the 4790K?
Did I stroke off and miss it or something?

 

[Idontcare]

Come to think of it, where the heck is AT's review of the 4790K?
Did I stroke off and miss it or something?

I think it is coming, just not published yet.

 

[WhoBeDaPlaya]

I think it is coming, just not published yet.

Cool, thanks for the info. Thought it was odd that I couldn't find a trace of the article.

 

[StinkyPinky]

Has anyone been able to test and quantify the thermal improvements on these Devil's Canyon chips like the 4790K? Supposedly they had addressed the TIM shortcomings with a new "polymer" TIM substance.

Mine idles at around 22C and hits around 60C playing Elder Scrolls Online at stock. This is on a crappy Arctic Freezer 7 Pro r2 cooler. I have no idea of ESO uses all the cores though.

I don't think this cooler would handle 4.7 but it should take 4.5 easily enough. I may try someday.

 

[BFG10K]

Mine idles at around 22C and hits around 60C playing Elder Scrolls Online at stock.

22C idle? I find that hard to believe. At what ambient?

 

[Majic 7]

Got mine together last night. Runs a little quieter than the 2600K and temps are pretty good so far. 29/30 and AfterBurner is reporting 24 on three of the cores out of the eight reported. The ASUS utility seems to be right around 30. Waiting for Steam to redownload a game and then I will check it under a load. Ended up nuking everything so I could install a new full version of 8.1, had an updated version before and new full seemed safer. I also have a 329 version of the CPU from MicroCenter. Downloaded Prime 95 and ran it for a few minutes and Afterburner is reporting 62 on all cores and the ASUS utility is reporting 38/40. Wasn't expecting that. This is at stock.

 

[StinkyPinky]

22C idle? I find that hard to believe. At what ambient?

PC has been on for hours. Don't know ambient but it's not cold.

My cooler sucks so it doesn't do so well under load.

 

[BFG10K]

It's either freezing in your room or that idle temperature reading is wrong. What does the BIOS temperature reading say?