My i7 7700k Has Arrived - Insights, Benchies & Overclocks Inside (Now With Delid!)

[RichUK]

I'm gonna have to delid my skylake so it doesn't feel so inadequate now! Thanks a lot Rich, it's all your fault.

I was hoping I wouldn't need to do it, but glad i did. Without delid, the 7700k's performance is poor from an OC perspective / expectation - delid unlocks the full potential. I have never delided a processor before, but it was dead simple - I just bought the tool and followed their video guide to delid and relid. I was hoping for a 20 degree drop, but happy with 30!

 

[thesmokingman]

Yea, that is an impressive drop. Did you notice any effect is voltage?

 

[Phynaz]

Prime is absolute crap, ive had an overclock for long time, using it for everything, benchmarking, gaming, and any other kind of stuff...tried prime and it failed instantly...

But i dont care about prime stability since ive never had a blue screen before using this prime crap..

So dont worry about prime stability..

If you aren't prime stable, you aren't stable. By definition.

 

[RichUK]

Yea, that is an impressive drop. Did you notice any effect is voltage?

No difference on voltage.

 

[MTDEW]

I was hoping I wouldn't need to do it, but glad i did. Without delid, the 7700k's performance is poor from an OC perspective / expectation - delid unlocks the full potential. I have never delided a processor before, but it was dead simple - I just bought the tool and followed their video guide to delid and relid. I was hoping for a 20 degree drop, but happy with 30!

I'm curious, did you use super glue when relidding like Rockit Cool shows in their video?
https://www.youtube.com/watch?v=bwg0HRw17lY
I bought the delid and relid kit, but I admit i'm a little hesitant to use super glue in case it needs to be redone in the future or if I dont get the CLU spread completely even the first try.

 

[Rifter]

If you aren't prime stable, you aren't stable. By definition.

Agreed

 

[RichUK]

I'm curious, did you use super glue when relidding like Rockit Cool shows in their video?
https://www.youtube.com/watch?v=bwg0HRw17lY
I bought the delid and relid kit, but I admit i'm a little hesitant to use super glue in case it needs to be redone in the future or if I dont get the CLU spread completely even the first try.

I did use super glue to relid (remember to let the glue set for a couple of hours, use in the corners). I used the same super glue and tape as recommended.

The tool is very precise, so if you're careful and diligent you shouldn't have any issues. It's so simple. The relidding tool also pushes down on the IHS to ensure a good contact.

Edit: I used a credit card to remove the black seal. Also remember to point the gold arrow on the processor in line with the arrow on the tool for correct orientation.

 

[RichUK]

Hi,
Fantastic findings RichUK, could you run Geekbench 4 on your latest 5.1 delided achievement ?
The 5Ghz (is it ?) GB you posted at 5879 single core seems disapointing

Geekbench at 5.3Ghz: https://browser.geekbench.com/v4/cpu/1357306

 

[cytg111]

Am I summing this up correctly? You had a "stock" stable overclock at 4.7, impressive enough in its own right, delidded at got a whooping 26 degrees average temp drop and are now honing in on a stable 5.0?
26 degrees yields about 300MHz?
Unless it throttles at stock-stock I dont see intel 'fumbling' with the TIM again..
(they should have named this one "The read Devils Canyon" and given it solder though.)

 

[vissarix]

Can

Geekbench at 5.3Ghz: https://browser.geekbench.com/v4/cpu/1357306

Can you do a Cinebench 11.5 single core and multicore run?

 

[Chrisch]

Geekbench at 5.3Ghz: https://browser.geekbench.com/v4/cpu/1357306

Your score seems low for 5.3GHz. This is what i get with only 5GHz

https://browser.geekbench.com/v4/cpu/1365712

Shut up and run the ryzen graphics file in blender...

(Just kidding with the shut up, but pretty pretty please? )
http://www.amd.com/en-us/innovations/new-horizon
directly under the video

 

[RichUK]

Your score seems low for 5.3GHz. This is what i get with only 5GHz

https://browser.geekbench.com/v4/cpu/1365712

It could be your memory speed being better than mine, as it's either my board or memory that's preventing higher clocks speeds even though they're rated at 3000mhz C15.

Plus you've boosted your uncore.

What temps are you getting under prime?

 

[Chrisch]

~60°C CoreTemp @ 5GHz but this CPU is also delidded, before delidding ~96-98°C with watercooling @ 5GHz

 

[RichUK]

~60°C CoreTemp @ 5GHz but this CPU is also delidded, before delidding ~96-98°C with watercooling @ 5GHz

It's good to see the temps aren't an isolated case and this is becoming a trend.

What batch number is yours? It should be the long alpha / numeric at the bottom of the heat spreader? Be interesting to see if a different batch also has the same issue as mine, as this will then be more indicative of what to expect come official release.

Also, for my sake, can you run a 5ghz Geekbench with everything else absolutely stock and set memory timings to 15-17-17-35 2T and post screenshots? I can then do a direct comparison to understand if it is my memory?

Also, what board are you running? I need to run a fan over my VRMs as they get too hot when running at 5ghz and throttle back to 4.2ghz every couple of seconds, so I'll likely get an Asus Z270 TUF when they're released.

Cheers.

 

[Chrisch]

i got 2 7700K

L643G202 runs 5GHz with 1.312v
L643G213 runs 5GHz with 1.264v

Board is a ASRock Z170 OC Formula with Bios P7.30

i try Geekbench in a few minutes.

 

[TheELF]

If this is with 150 samples than that's really impressive 20% lower then zen at 3,4Ghz and with only half the cores.

 

[Chrisch]

it is with 150 samples

@ RichUK

5GHz DDR4-3200 15-17-17-35 2T

5GHz DDR4-2133 15-17-17-35 2T

 

[RichUK]

it is with 150 samples

@ RichUK

5GHz DDR4-3200 15-17-17-35 2T

5GHz DDR4-2133 15-17-17-35 2T

Cheers thanks, I'll do a compare shortly with mine at 2133, purely for my own sanity.

 

[RichUK]

There also appears to be a new BIOS for my board (I noticed you weren't using v7 on your asrock), so might give that a try and hopefully get better memory speeds / timings.

 

[RichUK]

it is with 150 samples

@ RichUK

5GHz DDR4-3200 15-17-17-35 2T

5GHz DDR4-2133 15-17-17-35 2T

Thanks, Geekbench scores are essentially the same which shows I can get more from a better board / memory, which i'll be looking into next year when Z270 gets released.

https://browser.geekbench.com/v4/cpu/1366613

 

[jpiniero]

Can either of you run the Ryzen benchmark with the AVX2 optimized build?

 

[RichUK]

Well since updating the BIOS, this is the best Geekbench i can do:

CPU - 5.3Ghz
Uncore - 5Ghz
RAM: DDR4-2933 15-17-17-35 2T

https://browser.geekbench.com/v4/cpu/1368317

 

[jpiniero]

Not bad... seems that the top scores that aren't fake or corrupted are using DDR4 at very high speed... possibly in the mid to high 4000s.

 

[RichUK]

Not bad... seems that the top scores that aren't fake or corrupted are using DDR4 at very high speed... possibly in the mid to high 4000s.

Yeah, memory speed seems to make a sizable difference in Geekbench.

I'll try the AVX blender run a bit later.

 

[Madpacket]

So in summary was Tom's results correct? You gain around 300Mhz over the i7-6700K with sane voltages applied? 6 - 7% increase or so?