Confirmation if this is what i should be going for

[zjadow]

Right, so i need to upgrade my gaming rig (which i also want to run some (virtual) server 2012, hence why i picked 2 kits of 8GB -would like to know opinions if thats overkill or not)) for some courses i'm running. Parts need to be able to be delivered in Holland, hence why i picked the site below.
Still got 2 hdd's, 1 for storage, 1 SSD (240GB) and a coolermaster 600w SilentPro PSU which still runs fine, (does the PSU need an upgrade for w/e reason in your opinion? tell me! ), and an optical dvd drive, which i barely ever use.

Running 2x24" - 1 for gaming, 1 for sidestuff - so only big performance needed up to about 1920x1080, not going for mega resolutions.

CPU: http://azerty.nl/0-5654-620113/intel...-3-4-ghz-.html
Cooler: http://azerty.nl/0-976-458903/cooler...per-412s-.html
Memory: http://azerty.nl/0-750-478347/crucia...ix-sport-.html (x2)
GFX: http://azerty.nl/0-3921-482843/gigab...97oc-3gd-.html
Case: http://azerty.nl/0-1044-475292/corsa...ies-300r-.html
Motherboard: http://azerty.nl/0-5602-626119/gigabyte-z87m-d3h-.html

Any feedback is greatly appreciated.

[Deleted]

Major things to change are to go for an ATX form factor motherboard, and to change the cooler for either the Noctua D14 or the Corsair H60 2013 edition. Memory is fine (though I'd look elsewhere for 2*8GB). GPU could possibly be changed for this GTX770.

[zjadow]

Allright, yeah the motherboard i was going for initially was not in stock, so swapped it out - just looked at specs and missed the micro-atx part, my bad. Any suggestions there?

Noctua D14 and Corsair H60 don't seem to be suited for socket 1150? Or am i missing something?

GPU i agree on, it's a "small" increase in price and a lot of gain according to benchmarks, swapping that out, thx!

So, for 2x8gb - you'd get a different brand or what am i looking for?

Thx for the input so far : )

[Deleted]

Mobo: http://azerty.nl/8-5602-639716/msi-z...moederbor.html - Z87 G43 from MSI.

I explicitly mentioned the 2013 H60 water cooler as that is Haswell certified.

RAM is mostly personal preference, though that site seems really expensive for it, I'd suggest something like These Kingston sticks.

[glo]

Major things to change are to go for an ATX form factor motherboard, and to change the cooler for either the Noctua D14 or the Corsair H60 2013 edition. Memory is fine (though I'd look elsewhere for 2*8GB). GPU could possibly be changed for this GTX770.

H60 is pretty awful for the price, 25$ air coolers perform better unfortunately.

[Toffie]

H60 is pretty awful for the price, 25$ air coolers perform better unfortunately.

wat? Considering it's on par with the D14 I doubt that dude.
Benchmark : http://www.hitechlegion.com/reviews/...owall=&start=2

[glo]

wat? Considering it's on par with the D14 I doubt that dude.
Benchmark : http://www.hitechlegion.com/reviews/...owall=&start=2

That site's testing methodology is horrible, they don't even use temperature over delta. Regardless, let's look at one of the most reputable:

http://www.frostytech.com/articlevie...id=2705&page=6

They actually insert a temperature probe into a block of copper the size of an actual 1155/1150/2011 IHS while delivering so many watts through it. Anyway, see how the new H60 is hanging around the likes of the Gaia and H212+? It's nothing spectacular, and it certainly couldn't handle any kind of Haswell overclocking past 4.3. Hell, my H80i is capping out around 77c with idles in the mid 30s at 4.2/1.14v during AIDA64.

The only logical use of an H60 aside from looking cool is if you're going for a AIO cooler + blower style GPU for a full ambient air supply type setup.

[Toffie]

I can get my 4770k at 4.8 GHz at 1.37v without going over 90c under 10 runs of IBT, running a H60 with a quiet fan setup.
If you have your chip running at 1.14v with 77 under stress you either have a VERY bad soldered chip or you're not setting your H80 pump to the max.

Here is another test showing the H60 at 2-4 C behind the h100i which may be the best CPU cooler other than custom loops on the market.
http://www.guru3d.com/articles_pages...review,10.html

[glo]

I can get my 4770k at 4.8 GHz at 1.37v without going over 90c under 10 runs of IBT, running a H60 with a quiet fan setup.
If you have your chip running at 1.14v with 77 under stress you either have a VERY bad soldered chip or you're not setting your H80 pump to the max.

Here is another test showing the H60 at 2-4 C behind the h100i which may be the best CPU cooler other than custom loops on the market.
http://www.guru3d.com/articles_pages...review,10.html

I'm sorry, but the entire circle of review sites disagrees with that statement. Short of LN2 cooling, no one is running 4.8Ghz @ 1.37v. Go ahead and run AIDA64 and show us your temps, otherwise that's pure nonsense. Not to mention anything over 1.3v is bad for your chip, let alone 1.37...

[Toffie]

I'm sorry, but the entire circle of review sites disagrees with that statement. Short of LN2 cooling, no one is running 4.8Ghz @ 1.37v. Go ahead and run AIDA64 and show us your temps, otherwise that's pure nonsense. Not to mention anything over 1.3v is bad for your chip, let alone 1.37...

http://www.youtube.com/watch?v=WugDJbtBwwQ

Couple of chips hit 4.8GHz at 1.28-1.3v.

Now I don't have time to test the setup, but I will show my 3570K clock at almost 1.4V running lower than 80c under 8 hours of prime95 with a 212Evo...
http://www.mmo-champion.com/threads/...1#post20632036

Also delidding the chip can result in 20-30c less heat output if done right, all chips are different.

[glo]

http://www.youtube.com/watch?v=WugDJbtBwwQ

Couple of chips hit 4.8GHz at 1.28-1.3v.

Now I don't have time to test the setup, but I will show my 3570K clock at almost 1.4V running lower than 80c under 8 hours of prime95 with a 212Evo...
http://www.mmo-champion.com/threads/...1#post20632036

Also delidding the chip can result in 20-30c less heat output if done right, all chips are different.

We're not talking about Ivy Bridge, thermals and voltages are entirely different from Haswell. He's not using an AVX stress test, he says it right before he gets into the benches. Additionally his voltages don't go anywhere close to your stated 1.37v. Lastly, he's using an open test bench, not an enclosed case with his temps spiking to 100c frequently. What do you think those temperatures would be in an enclosure?

Just like it's been discussed everywhere, you need a perfect chip to go above 4.6 with acceptable thermals, most aren't perfect.

Listen to the end of his video where he goes on to say he got super lucky with most of those chips coming from the same wafer, one being an engineering sample and the one random chip he had laying around performing poorly.

P.S. - You said you were getting 4.8Ghz on your 4770k at 1.37v on an H60. He had a perfect batch, was using an H100i and was capping out at 100c on most chips well under 1.3v. Do you have a magical H60 that performs better in an enclosure than his open test bench running an H100i?

[Fluorescent0]

Did you by any chance delid?

[Toffie]

You don't need a perfect chip to hit 4.8GHz, you need a average chip and then delid it which frees room for some extra overclocking.
Like I said, all chips are different in terms of performance and thermals. You might get a 4770k that runs 20c higher than other 4770k's at the same voltage.

Also temps are pretty much close to Ivy Bridge, but it's really random since the glue might be less in some chips helping the surface to transfer better.
http://www.techpowerup.com/reviews/I...parison/9.html

Did you by any chance delid?

I have with my 4770k, not the 3570k I had earlier which was running amazing temps. I think the temps decreased by about 20-25c at the hottest core which freed up my 1.2v limit.

[glo]

You don't need a perfect chip to hit 4.8GHz, you need a average chip and then delid it which frees room for some extra overclocking.
Like I said, all chips are different in terms of performance and thermals. You might get a 4770k that runs 20c higher than other 4770k's at the same voltage.

Also temps are pretty much close to Ivy Bridge, but it's really random since the glue might be less in some chips helping the surface to transfer better.
http://www.techpowerup.com/reviews/I...parison/9.html




I have with my 4770k, not the 3570k I had earlier which was running amazing temps. I think the temps decreased by about 20-25c at the hottest core which freed up my 1.2v limit.

Delidding doesn't magically make a chip require less voltage to be stable at a certain clock, nor did you mention delidding in the first place. Either way, delidded or not you're not pushing 1.37v through a 4.8 overclock on Haswell.

Further, temps are no where close at higher clocks. You're linking a review saying that they're close that only show the differences between Haswell and nothing else. They're also using a full custom liquid loop.

[Toffie]

I quote "Haswell temperatures are equal to those of Ivy Bridge processors. Temperatures may sometimes be higher due to the integrated VRM inside the CPU (iVR)."
Again "Yeah, I have tested 6x i7 4770K in total, these samples had really bad TIM (not that the retails should be that better though)."

He is pushing 78 c with a custom loop at 1.24v (4.7GHz) without any delidding. With delidding he could be running 50 c under load, 60-65c with 1.37v which is 25-30c less than what my H60 achieves.

Anyway, you are hard to reason with, If you don't believe my claims then I have nothing more to say.

- - - Updated - - -

Delidding doesn't magically make a chip require less voltage to be stable at a certain clock

No one said that, I was merely stating that it lifted the 1.2v I had which was running wild temps and the delidding freed up my voltage limited by the temps.

- I have linked you proof of my earlier Ivy Bridge running close to 1.4v at 75c under load with the 212evo, please don't tell me you think Haswell runs 25c+ hotter..

[Faithh]

Hmm a H60 or H80 or H100 is just totally useless for LGA115x except for looks. The backplates just suck, it's not making proper contact at all compared on S2011 where the mounting is complety different and much more secure. Pushing my old 2600K to 120-130Watts (Coretemp) was giving me 90°+ and on the other hand on my 3930K 200Watts (Coretemp) I was able to keep it below 80°. No wonder the rad never felt hot anyways on lga 1155.

I really would advise you a NH-U14s which performs very close to a NH-D14 and a lot of less noise than AIO's. See here the testresults of all the coolers: http://be.hardware.info/productinfo/...elers?products[]=183458&specId=10791&tcId=312

It's really cheap for what you can especially price/performance/noise ratio 63eur here -> http://afuture.nl/productview.php?productID=968245

[zjadow]

Thanks all for the input so far. The noctua NH-U14s seems to perform really well indeed, will update that.

Now for an updated list:
Mobo (Azerty couldn't deliver for next few days on G43 or G45 gaming series, so picked a different site for this):
http://www.alternate.nl/MSI/MSI+Z87-...oduct/1080880/
CPU: http://azerty.nl/0-5654-620113/intel...-3-4-ghz-.html
GFX: http://azerty.nl/0-5644-626550/msi-gtx-770-gaming-.html
Cooler: http://azerty.nl/0-976-619238/noctua-nh-u14s-.html
Case: http://azerty.nl/0-1044-475292/corsa...ies-300r-.html
Ram: http://azerty.nl/0-750-602014/kingst...y-series-.html

Any last comments on this before i'd order?

[Toffie]

Everything looks good, I would pick some different colored ram for your red/black setup though.

[zjadow]

Agreed, swapped to a more red/black-ish kit
Thanks for the help to all who replied, topic closed for as far as i'm concerned for this build.