• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Finished My Overclock. How does it all look? 4.9ghz @ 1.42 vcore.

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

GenieGOR

Member
Joined
Jan 9, 2016
So I have set my Base clock @ 111, and multiplier to 44. I get 4.88 ghz. I am tipping right at 4.9. I may do 112 base to get there. My vcore is set to 1.4 in the bios, if I go any lower it will crash and be unstable. I also see that even tho I put it at 1.4 it jumps to 1.42 under gaming or CPUZ single core bench. Under multi-core CPUZ bench it jumps to 1.44. I also had to set the specs below:

CPU Vcore - 1.4v
system agent voltage - 1.300v
CPU CIO - 1.250v

Here are my aida, and cpuz benchmarks.

How is everything looking guys as far as speed and performance? Anything I should change?
 

Attachments

  • CPUZ - Bench - info - multicore.JPG
    CPUZ - Bench - info - multicore.JPG
    63.9 KB · Views: 857
  • CPUZ - Bench - info.JPG
    CPUZ - Bench - info.JPG
    63.9 KB · Views: 988
  • CPUZ - Bench.JPG
    CPUZ - Bench.JPG
    42.2 KB · Views: 869
  • GPU Test 3.png
    GPU Test 3.png
    358.8 KB · Views: 921
  • Memory test 2.png
    Memory test 2.png
    201.7 KB · Views: 1,000

Johan45

Benching Team Leader Super Moderator
Joined
Dec 19, 2012
Looks nice but none of those are stability tests. You need to use OCCT, IBT or AIDA64 stability tests for testing temps and stability
 
OP
G

GenieGOR

Member
Joined
Jan 9, 2016
Looks nice but none of those are stability tests. You need to use OCCT, IBT or AIDA64 stability tests for testing temps and stability

I was just asking as far as voltages, and speed. If those set voltages I had were good and safe and if the speed its producing are good results.

EDIT: FSB, and memory stress fail. I turned down the base clock to 109, and the cpu vcore to 1.4. I am now doing the stress with no fail on AIDA. So its now running @ 1.45 vcore and 4.79 ghz.
 

Attachments

  • stabilitytest1.png
    stabilitytest1.png
    23 KB · Views: 800
  • stabilitytest.png
    stabilitytest.png
    31.3 KB · Views: 815
  • stabilitytest3.png
    stabilitytest3.png
    48.7 KB · Views: 813
Last edited:

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
Well... we said to stay under 1.42V and temps under 90C when stress testing. If you are at those values or less, you are good to go.

Since you are still failing, I would lower the clocks a little bit and voltages.
 
OP
G

GenieGOR

Member
Joined
Jan 9, 2016
Well... we said to stay under 1.42V and temps under 90C when stress testing. If you are at those values or less, you are good to go.

Since you are still failing, I would lower the clocks a little bit and voltages.

Thanks Earthdog. It seems CPUZ is not reliable it shows 1.45 under vcore but AIDA shows 1.192. Not sure who to trust. Also what would be faster DDR4 18-18-18-40 timings @ 3480 MHz or DDR4 9-9-9-19 @ 2100 MHz?
 
OP
G

GenieGOR

Member
Joined
Jan 9, 2016
Also memory speed and L2 cache speeds increased after bringing down voltage a bit and decreasing the base to 109.
 

Attachments

  • Memory test 2.png
    Memory test 2.png
    201.7 KB · Views: 895
  • Memory test 3.png
    Memory test 3.png
    201.5 KB · Views: 884

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
Id trust the voltage readings off the board (you need a multimeter) and nothing else. What do you have it set to in the BIOS would be important to know!

As far as your memory question, I don't see a point in buying DDR4 ram with DDR3 speeds (2100MHz).

You can run AIDA back to back to back with no changes and get different results. ;)
 
OP
G

GenieGOR

Member
Joined
Jan 9, 2016
Id trust the voltage readings off the board (you need a multimeter) and nothing else. What do you have it set to in the BIOS would be important to know!

As far as your memory question, I don't see a point in buying DDR4 ram with DDR3 speeds (2100MHz).

You can run AIDA back to back to back with no changes and get different results. ;)

I was basically asking if 18-18-18-40 timings were actually trading off the speed, making it slower. I have another machine here running 9-9-9-19 timings. DDR3. So I was wondering. Also the cpu vcore is set to 1.4 in the bios. It still goes over tho.
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
Also the cpu vcore is set to 1.4 in the bios. It still goes over tho.
Again, the only way to know for sure is to use a MM on the voltage read points on that board. There is a reason you paid a premium for that board... you should use that feature... ;)
 

JrClocker

AKA: JrMiyagi
Joined
Sep 25, 2015
If you are shooting for 4.8 or 4.9 GHz...have you tried lowering BCLK to 100 MHz and using a 48 or 49 multiplier?

Less moving parts (i.e. you are not overclocking the memory, cache, and everything else that runs off the BCLK).

You didn't say anything about your cache...is that still at default? Increasing the BCLK by 10% will also increase the cache frequency...could cause instabilities if you have not tweaked there yet.
 
OP
G

GenieGOR

Member
Joined
Jan 9, 2016
If you are shooting for 4.8 or 4.9 GHz...have you tried lowering BCLK to 100 MHz and using a 48 or 49 multiplier?

Less moving parts (i.e. you are not overclocking the memory, cache, and everything else that runs off the BCLK).

You didn't say anything about your cache...is that still at default? Increasing the BCLK by 10% will also increase the cache frequency...could cause instabilities if you have not tweaked there yet.

Yes I tweaked cache as well. How would I go about overclocking the ram and cpu and leaving cache stable?
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
I honestly wouldn't bother overclocking either as both return very little performance gains.

That said, you simply need to up the mulitplier for both and leave bclk alone (course adjustments). Or use BCLK and lower CPU multiplier for finer control.

Again, I wouldn't bother. The time and effort it takes versus what you get out of it, isn't worth the time.
 

Johan45

Benching Team Leader Super Moderator
Joined
Dec 19, 2012
My suggestion would be to start from ground zero . Hit optimized default in BIOS. Set XMP profile. Set the Voltage to 1.35. Start raising the multiplier and test in between keeping the temps in check
 
OP
G

GenieGOR

Member
Joined
Jan 9, 2016
Right now I have a very stable 4.79ghz @ 1.42-1.44 vcore. It is set to 1.380 in bios but still goes to 1.42-1.44.

I also have my bclk to 109 now and multiplier to 44. My CIO voltage to 1.21 and system agent to 1.26.

Xmp on and memory at 3466.

It's running 25,650 in firestrike, and runs Aida stress for a hour no issues.

Are these safe values?
 
Last edited:

Johan45

Benching Team Leader Super Moderator
Joined
Dec 19, 2012
I think for daily use I'd try to get my V_Core down around the 1.35v mark. 1.44v is kind of high IMO
 

JrClocker

AKA: JrMiyagi
Joined
Sep 25, 2015
Agree with these folks.

Also, you screen shots showed 86 C max on Core 2 during stress testing.

You never mentioned what type of cooling you are using.

Also, it's winter (at least in the Northern Hemisphere). 86 C max for a daily overclock during winter might make things a toasty when it heats up in a few months.

I live in Florida...this time of year my office where my PCs sit can get in the 29 C to 30 C range (heat is on...no A/C). In the summer, it can get up to 34 C during the day (while my A/C is turned down as no people are home.)

Might be good to shoot for max temps under 80 C...and give yourself some headroom.
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
That is also a stress testing temperature...most load temps (say gaming, etc) are a lot less. ;)