- Feb 14, 2005
- 10,341
- 678
- 126
2nd January - 5.1Ghz Prime Stable with AVX multiplier offset
Interesting, so the new BIOS actually allows a negative multiplier offset for AVX instructions which has allowed me to overclock to 5.1Ghz and auto drop down to 5Ghz when avx instructions are executed. Not sure if this is a new feature for Kaby Lake processors?
I have left the voltage configuration the same as the 5Ghz overclock that was 24 hour Prime95 v27.9 stable - 230mv offset with LLC level 2 - and now have prime95 v26.6 stability at 5.1Ghz for over 2 hours using the 1344 small FFT custom test (guide here: http://overclocking.guide/stability-testing-with-prime-95/). The voltage stabilises at 1.376v (instead of 1.36v when using v27.9 @ 5Ghz) as v26.6 isn't putting as much load on the processor.
When I execute Prime95 v27.9 or v28.10 the multiple automatically drops to x50 / 5Ghz as expected, so I have essentially found another 100Mhz using the AVX offset.
Some 5.3Ghz Cinebench R15, why not:
~~~~~~~~~~~~~~
21st December - 5Ghz 24 hours Prime Stable
Goal Achieved
I set out to achieve 5Ghz at or below 1.4v and 75 degrees.
Associating benchmarks:
Next Steps:
~~~~~~~~~~~~~
20th December
Member Chrisch posted some great results from his gold sample 7700k - 5Ghz @ 1.28v and 5.1Ghz @ 1.328v under Prime is extremely impressive.
When delidded the max temps in Prime are only ~50 degrees at 5Ghz on this chip (made possible by the reduced Vcore requirement for 5Ghz), netting a ~20 degree reduction in max temps. Let's hope more 7700k's perform like batch L642G244!
5Ghz @ 1.28v
5.1Ghz @ 1.328v
Before & after delidding, 30min 1344k Prime test:
Chrisch's delid video - it really is that simple (and relatively risk free IMO)!
~~~~~~~~~~~~~~~~~~~
18th December
Member Chrisch has also witnessed similar temp drops after a delid on a 7700k with a different batch number, so this appears to be a trend (albeit with only 2 data points).
Chrisch also posted an impressive Geekbench score - 5GHz, DDR4-3200 15-17-17-35 2T:
~~~~~~~~~~~
17th December
Geekbench at 5.3Ghz / memory stock: https://browser.geekbench.com/v4/cpu/1357306
~~~~~~~~~~~
16th December
EDIT:
Out of curiosity I googled 7700k delid and saw a number of websites reporting on my findings of the 30 degree temp drop, but were unsure on cooling method in both tests so thought I should clarify: The before and after tests were 100% identical (with exception of delid) to isolate the delid, both tests used:
The journey to rock solid 5Ghz continues.
Some more interesting finds. Before delid the temps in Prime 95 v28.7 Small FFTs were insane and would almost instantly fail - i believe the increased temps destabilises overclocks (from what I've read, due to noise caused by heat).
After delid, I was able to run Prime 95 v28.10 small FFTs (latest version!) at 5Ghz 1.392v for over 7 hours before one logical core failed.
At 5Ghz using Prime 95 v28.10 Small FFTs, I observed:
Ryzen blender test at 5.2Ghz:
~~~~~~~~~~~~~~~~~~~~~~
13th December
EDIT: Initial Intel Burn Test shows a pass at 5.1Ghz - 1.376v / 72 degrees max temp: https://s27.postimg.org/8fewhhwkz/IBT_5_1_Ghz.jpg
Delid done
Before and after delid with Liquid Ultra - 5Ghz, 1.344v, Kraken X62 @ 50% fan and 65% pump (i.e. silent), Prime95 v27.9 - 26 degrees difference (~30 degrees max temp difference)!
Before - https://s27.postimg.org/mgu5i6sc3/5ghz_P95_27_9_before.jpg
After - https://s28.postimg.org/lg9hfqj6l/5ghz_P95_27_9_after.jpg
Prime95 v28.7 doesn't instantly fail at 5Ghz anymore either. It would shoot up to 99 degrees, now stays at 66 degrees and did so for 1 minutes no failure.
More testing to follow!
https://s23.postimg.org/ko320iaiz/image.jpg
https://s24.postimg.org/v8m16c8n9/image.jpg
https://s30.postimg.org/97zrj5fgh/image.jpg
https://s23.postimg.org/c91oadnnf/image.jpg
https://s29.postimg.org/50d40oth3/image.jpg
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
6th December
EDIT: - Looks like 5Ghz might be back on the table
Short test with Prime95 v27.9 - 5Ghz at 1.34v stable for 15 minutes where as v28.7 was insta-fail and insane temps:
For note, I am also not seeing failures when running IBT and OCCT (AVX Linpak ticked) at 5Ghz.
--------------------Everything below this line is old case and old H110i cooler---------------
So, first stability checkpoint / baseline reached. The headlines:
Given i'll need more voltage for higher frequencies resulting in 99 Degrees temps on the current setup (where it will also continually throttle), I'll be waiting until I delid and install the Kraken X62 on Thursday/Friday before going for Prime 95 Small FFT stability at 4.9Ghz which should be achievable with less than 1.4v (i may even attempt 5Ghz again, with better VRM cooling and CPU temps).
Benchmark sanity test:
The below should be a useful comparison against 6700k overclocks:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1st December
It's too early to draw conclusions, but my initial observations are:
Next Steps:
30th November
Right, let's get to it (i'll tidy this up later, along with adding further detail regarding the OC (and Temps
)
Rig:
Proc - i7 7700k
Mobo - Asrock Z170 Pro4S - BIOS v7
RAM - Corsair 2x8GB 3000Mhz C15
GFX - EVGA GTX 1060 6GB SC
Cooling Stock - an old Thermalright 6 heat pipe tower, Corsair SP120
Cooling for OC - Corsair H110i with 2 x Corsair ML140 Pro LED (cheap black Friday deal)
Processor Pics:
https://s18.postimg.org/42y5ihjrt/image.jpg
https://s15.postimg.org/5gwhnw2nf/image.jpg
https://s11.postimg.org/f98al9mgz/image.jpg
GeekBench v4: https://browser.geekbench.com/v4/cpu/1203097
CPU-Z (Stock):
Cinebench R15 (Stock - Single & Multi Threaded Scores):
Interesting, so the new BIOS actually allows a negative multiplier offset for AVX instructions which has allowed me to overclock to 5.1Ghz and auto drop down to 5Ghz when avx instructions are executed. Not sure if this is a new feature for Kaby Lake processors?
I have left the voltage configuration the same as the 5Ghz overclock that was 24 hour Prime95 v27.9 stable - 230mv offset with LLC level 2 - and now have prime95 v26.6 stability at 5.1Ghz for over 2 hours using the 1344 small FFT custom test (guide here: http://overclocking.guide/stability-testing-with-prime-95/). The voltage stabilises at 1.376v (instead of 1.36v when using v27.9 @ 5Ghz) as v26.6 isn't putting as much load on the processor.
When I execute Prime95 v27.9 or v28.10 the multiple automatically drops to x50 / 5Ghz as expected, so I have essentially found another 100Mhz using the AVX offset.


Some 5.3Ghz Cinebench R15, why not:

~~~~~~~~~~~~~~
21st December - 5Ghz 24 hours Prime Stable
Goal Achieved
I set out to achieve 5Ghz at or below 1.4v and 75 degrees.
- 5Ghz @ 1.36v
- Average temps under 75 degrees
- VCCIO / VCCSA - Auto in BIOS (default)
- 24 hours Prime95 v27.9 Small FFTs - stable with no failures
- I am comfortable with v27.9 over v28.10 (more volts / temps) to prove stability for my usage.
- I reduced the Kraken x62 fan speed to 50% and increased pump to 100% - the system is now silent for me.

Associating benchmarks:
- Geekbench: https://browser.geekbench.com/v4/cpu/1393935
- Cinebench R15: https://s28.postimg.org/uo1q2uzv1/CB_R15.jpg
- Cinebench R11.5: https://s24.postimg.org/uu5mqe9z9/CB_R11_5.jpg
- Blender v2.77 - Ryzen @150 Samples: https://s28.postimg.org/5kxlq8ny5/Blender_V2_77_Ryzen.jpg
- Blender v2.78a - Ryzen @150 Samples: https://s29.postimg.org/rqye5er9j/Blender_V2_78a_Ryzen.jpg
Next Steps:
- Z270 (Asus Hero or TUF) and DDR4-4000+
~~~~~~~~~~~~~
20th December
Member Chrisch posted some great results from his gold sample 7700k - 5Ghz @ 1.28v and 5.1Ghz @ 1.328v under Prime is extremely impressive.
When delidded the max temps in Prime are only ~50 degrees at 5Ghz on this chip (made possible by the reduced Vcore requirement for 5Ghz), netting a ~20 degree reduction in max temps. Let's hope more 7700k's perform like batch L642G244!
5Ghz @ 1.28v

5.1Ghz @ 1.328v

Before & after delidding, 30min 1344k Prime test:

Chrisch's delid video - it really is that simple (and relatively risk free IMO)!
~~~~~~~~~~~~~~~~~~~
18th December
Member Chrisch has also witnessed similar temp drops after a delid on a 7700k with a different batch number, so this appears to be a trend (albeit with only 2 data points).
Chrisch also posted an impressive Geekbench score - 5GHz, DDR4-3200 15-17-17-35 2T:

~~~~~~~~~~~
17th December
Geekbench at 5.3Ghz / memory stock: https://browser.geekbench.com/v4/cpu/1357306
~~~~~~~~~~~
16th December
EDIT:
Out of curiosity I googled 7700k delid and saw a number of websites reporting on my findings of the 30 degree temp drop, but were unsure on cooling method in both tests so thought I should clarify: The before and after tests were 100% identical (with exception of delid) to isolate the delid, both tests used:
- The same cooler at the same setting: Kraken X62 - Fan @ 50% and pump at @ 65% (by the way, this cooler is only about 3 degrees cooler than the H110i, but has better config available)
- Exactly the same BIOS settings - which was a x50 multiplier and voltage offset with LLC level 2 resulting in a stable 1.344v under prime 95 load
- Memory stock (C15) and Uncore stock.
- Prime 95 v27.9 run for 15 minutes in both tests.
- The same case / config / install (tested with case side panel installed - i.e. how i would actually use it for 24/7 use so the test would be more meaningful to me): https://s30.postimg.org/gzluzd7w1/Case.jpg
The journey to rock solid 5Ghz continues.
Some more interesting finds. Before delid the temps in Prime 95 v28.7 Small FFTs were insane and would almost instantly fail - i believe the increased temps destabilises overclocks (from what I've read, due to noise caused by heat).
After delid, I was able to run Prime 95 v28.10 small FFTs (latest version!) at 5Ghz 1.392v for over 7 hours before one logical core failed.
At 5Ghz using Prime 95 v28.10 Small FFTs, I observed:
- 1.36v fails in 9 mins (one logical core)
- 1.376v fails in 17 mins (one logical core)
- 1.392v fails on one logical core in 7 hours 18 minutes.

Ryzen blender test at 5.2Ghz:

~~~~~~~~~~~~~~~~~~~~~~
13th December
EDIT: Initial Intel Burn Test shows a pass at 5.1Ghz - 1.376v / 72 degrees max temp: https://s27.postimg.org/8fewhhwkz/IBT_5_1_Ghz.jpg
Delid done
Before and after delid with Liquid Ultra - 5Ghz, 1.344v, Kraken X62 @ 50% fan and 65% pump (i.e. silent), Prime95 v27.9 - 26 degrees difference (~30 degrees max temp difference)!
Before - https://s27.postimg.org/mgu5i6sc3/5ghz_P95_27_9_before.jpg
After - https://s28.postimg.org/lg9hfqj6l/5ghz_P95_27_9_after.jpg
Prime95 v28.7 doesn't instantly fail at 5Ghz anymore either. It would shoot up to 99 degrees, now stays at 66 degrees and did so for 1 minutes no failure.
More testing to follow!


https://s23.postimg.org/ko320iaiz/image.jpg
https://s24.postimg.org/v8m16c8n9/image.jpg
https://s30.postimg.org/97zrj5fgh/image.jpg
https://s23.postimg.org/c91oadnnf/image.jpg
https://s29.postimg.org/50d40oth3/image.jpg

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
6th December
EDIT: - Looks like 5Ghz might be back on the table
Short test with Prime95 v27.9 - 5Ghz at 1.34v stable for 15 minutes where as v28.7 was insta-fail and insane temps:
For note, I am also not seeing failures when running IBT and OCCT (AVX Linpak ticked) at 5Ghz.

--------------------Everything below this line is old case and old H110i cooler---------------
So, first stability checkpoint / baseline reached. The headlines:
- 6 hours Prime95 Small FFTs stable - using Prime95 v28.7
- 4.7 Ghz @ 1.264v (memory stock - 2133) - note, HWinfo doesn't report volts correctly - see CPU-z.
- Average temp 83 degrees
- Corsair H110i set to quiet mode (I'd never leave it in performance mode due to the increased noise - I don't want to sit next to a vacuum cleaner!)
- A-Tuning was used to adjust the offset - 0.02v and 0.03v failed in under an hour, 0.05v achieved stability at 4.7Ghz.
Given i'll need more voltage for higher frequencies resulting in 99 Degrees temps on the current setup (where it will also continually throttle), I'll be waiting until I delid and install the Kraken X62 on Thursday/Friday before going for Prime 95 Small FFT stability at 4.9Ghz which should be achievable with less than 1.4v (i may even attempt 5Ghz again, with better VRM cooling and CPU temps).
Benchmark sanity test:
The below should be a useful comparison against 6700k overclocks:

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1st December
It's too early to draw conclusions, but my initial observations are:
- The 7700k has the potential to be stable at 5Ghz @ ~1.4v, with exceptional samples reaching 5.1 / 5.2 perhaps.
- The power draw when overclocked 'seems' high, evidenced by the vdroop of 0.08v when under IBT load! I have no other means of measuring power draw - Let's hope the AT review focuses in on this.
- I am unclear at this point whether temps are being reported correctly, but i believe not given the general expectations is for higher temps under poor air cooling (plus I witnessed a discrepancy of ~15 degrees idle temp between the BIOS and Windows/OS) .
- Could an AIO cooler be standard in order to run overclocks north of 4.8ghz?
- If you're going to overclock, get a board that has strong / stable power delivery to sustain voltage with limited vdroop! I'm not sure whether Z270 motherboards will offer better VRMs etc in anticipation for the 7700k.
Next Steps:
- Install AIO water cooler (either NZXT Kraken X62 or Corsair H110i with better fans)
- Verify temperature readings
- Configure voltage to avoid it dropping below stable voltage requirements at 5Ghz (1.36 - 1.38v should be enough is my assumption)
- Stress test
- Re-run benchmarks
- De-lid and log temp drop - tool: http://rockitcool.myshopify.com/
- Enjoy
- Buy a decent Z270 board if needed when they're released.
- Enjoy some more.
30th November
Right, let's get to it (i'll tidy this up later, along with adding further detail regarding the OC (and Temps
Rig:
Proc - i7 7700k
Mobo - Asrock Z170 Pro4S - BIOS v7
RAM - Corsair 2x8GB 3000Mhz C15
GFX - EVGA GTX 1060 6GB SC
Cooling Stock - an old Thermalright 6 heat pipe tower, Corsair SP120
Cooling for OC - Corsair H110i with 2 x Corsair ML140 Pro LED (cheap black Friday deal)
Processor Pics:

https://s18.postimg.org/42y5ihjrt/image.jpg
https://s15.postimg.org/5gwhnw2nf/image.jpg
https://s11.postimg.org/f98al9mgz/image.jpg
GeekBench v4: https://browser.geekbench.com/v4/cpu/1203097
CPU-Z (Stock):

Cinebench R15 (Stock - Single & Multi Threaded Scores):

Last edited: