Asus Motherboard + 11700K = All Core Boost Under Advertised Intel Spec? Bad Chip?

[plastiscɧ]

I'll bet your chip can do 4.8GHz all core at the same voltage mine does 4.6GHz lol. I need a good chip like yours for my build.

I'll keep tweaking. Maybe I'll figure out the better cores (are they the cores with a * next to them in BIOS?) and give just them the 4.8GHz boost while the rest are 4.5GHz. IDK.

It is doing 50 all cores with slightly less voltage. i knew this but i was greedy. i will observe it and maybe lower all a bit more to comfort him more..

more like 51-51-50-50-49-49-48-48

 

[CyberCT]

without running the Realbench 2.56 for 8 hours, we don't know if your OC is 100% stable (even though probably stable for everything else) but you definitely do have a better chip than I do.

 

[plastiscɧ]

without running the Realbench 2.56 for 8 hours, we don't know if your OC is 100% stable (even though probably stable for everything else) but you definitely do have a better chip than I do.

okay. so this was just a shot. what u gonna say?

 

[CyberCT]

The stress test option to the right ... not the benchmark

 

[plastiscɧ]

The stress test option to the right ... not the benchmark

it passed the 15 minutes test. thanks for this suggestion. i don't think i need to run it for 8 hours, i'm too "good-natured" for this after all.
but i will keep an eye on those programs for the case changing something.

 

[CyberCT]

All of my voltage offset attempts passed through the first hour ... the failed tests started at roughtl 2+ hours in. I did these tests overnight so I can't tell the exact time of the first failure (no timestamps in this program).

 

[CyberCT]

Tried Realbench v2.43 instead and no errors for 8 hours. Ran 4 hours of Prime 95 small ffts and no errors. So with temps in check the most I can boost to is 4.6GHz each core. Since I won't build another PC for another @4 years, I decided to just buy an 11900K instead since it's just a binned 11700K so to speak. On my 11900K I can do 4.8GHz all core (actually just limiting boost to 4.8GHz per core) with -0.04 f/v offset per core that passes 8 hours Realbench. Highest core temp is 72C. However, there is a MemTest error detected after 200% coverage so I'll have to start tweaking down that f/v offset to find the right error free number. Unfortunately the 11700K is SP55 and the 11900K is SP50 according to the Asus BIOS.

Going to try and find the best Gear1 setting for the DDR4 4000 6-16-16-36 RAM after testing everything stock to make sure the RAM isn't faulty. Just started using MemTest on this build for stability testing.

CORSAIR Vengeance LPX 16GB (2 x 8GB) 288-Pin DDR4 SDRAM DDR4 4000 (PC4 32000) AMD Optimized Desktop Memory Model CMK16GX4M2Z4000C16 - Newegg.com

Buy CORSAIR Vengeance LPX 16GB (2 x 8GB) 288-Pin DDR4 SDRAM DDR4 4000 (PC4 32000) AMD Optimized Desktop Memory Model CMK16GX4M2Z4000C16 with fast shipping and top-rated customer service. Once you know, you Newegg!

www.newegg.com

 

[hymced]

@CyberCT sorry to spam your post, I have never done OCing, so I am not understanding everything here (yet! )
But I have a similar config to yours I think, so if you could ask me for specific tests/values I could take a look.
Maybe you can also help me and maybe in the last month you have found some things I haven't:

https://www.reddit.com/r/intel/comments/th0akd

 

[ThrashZone]

Hi,
Throttling is yes silly temperture limits but also silly timing limits too.

Digi power options max these out

Maximum CPU Core Temperature [115] this one can be anything you feel good about 100c...
CPU Core/Cache Current Limit Max. [255.75]
Long Duration Package Power Limit [4095]
Package Power Time Window [448]
Short Duration Package Power Limit [4095]

 

[hymced]

Hi,
Throttling is yes silly temperture limits but also silly timing limits too.

Digi power options max these out

Maximum CPU Core Temperature [115] this one can be anything you feel good about 100c...
CPU Core/Cache Current Limit Max. [255.75]
Long Duration Package Power Limit [4095]
Package Power Time Window [448]
Short Duration Package Power Limit [4095]

Hello TrashZone, are you talking about my case or @CyberCT's one?
As I mentioned in my post, I am quite a newbie... I don't understand your answer ^^ digi power?
I enabled XMP I timings, it is a bit better but not by much (100 pts), and CPU-Z bench is a bit worse, does it corroborate your timings hypothetis?
My goal is not necesseraly to reach the results I could find on the web, but rather to understand if they should be read as stock results (if they should, I am a bit short, and I'd like to know why), or as possibly OCed results (in this case I am sure there are a lot of ways to reach them easily with OCing techniques you master better than me )

 

[ThrashZone]

Hi,
Original poster but really anyone on the platform

Digi-power settings are asus mother board bios settings but I'm sure other board manufactures has similar options

Think the time limit is the worst and wouldn't be all that extreme of a change it would just stop the throttling until the chore was actually done.

Package Power Time Window [448]

I don't have an image of the digi-power bios page but it's listed in extreme tweaker settings on an asus board as an example.

 

[hymced]

Ok I also do have an ASUS mobo (Z590-I) and the Package Power Time Windows is set to Auto (56s), so its greater than the time R23 need for Pass 1 (my results are only for Pass 1 which last around 50s). So I am not sure maxing it out would help, don't you think?

 

[AusWolf]

I made a little "guide" on PL and TAU values with 11th gen, so that we're all on the same page.

Edit: Actually, PL2 can be higher than MCE depending on the motherboard. But the point stands.

Ok I also do have an ASUS mobo (Z590-I) and the Package Power Time Windows is set to Auto (56s), so its greater than the time R23 need for Pass 1 (my results are only for Pass 1 which last around 50s). So I am not sure maxing it out would help, don't you think?

That depends on what your goal is. If you want max. clocks at PL1 at all times, then yes. Though I think setting max. power limit values is a better way, because TAU will never expire then.

 

[hymced]

That depends on what your goal is. If you want max. clocks at PL1 at all times, then yes. Though I think setting max. power limit values is a better way, because TAU will never expire then.

I don't want anything really but to understand why my Cinebench R23 stock score is that low compared to online benchmark values
I did what @ThrashZone suggested, setting Package Power Time Window to 448s instead of default 5-s, it doesn't change a thing, at best I got 14850 pt

 

[eidairaman1]

Ymmv

 

[AusWolf]

I don't want anything really but to understand why my Cinebench R23 stock score is that low compared to online benchmark values
I did what @ThrashZone suggested, setting Package Power Time Window to 448s instead of default 5-s, it doesn't change a thing, at best I got 14850 pt

What's wrong with 14850? It's a decent score as far as I'm concerned.

 

[hymced]

What's wrong with 14850? It's a decent score as far as I'm concerned.

I have listed all the reference to cinebench R23 benchmark I could find with this cpu, and it seems mine is behind all of them. Somebody told me that all 11700Ks should give around the same STOCK score. All the values I found are in the same 1500+ range, so that corroborates the statement. But I think mine is not quite in that range... Don't you think that the difference is significant? I even saw an R23 scoring 15276 pt (stock I think, overclocking is not mentioned) here: https://www.tweaktown.com/reviews/9839/intel-core-i7-11700k-cpu/index.html

 

[Mussels]

I have listed all the reference to cinebench R23 benchmark I could find with this cpu, and it seems mine is behind all of them. Somebody told me that all 11700Ks should give around the same STOCK score. All the values I found are in the same 1500+ range, so that corroborates the statement. But I think mine is not quite in that range... Don't you think that the difference is significant? I even saw an R23 scoring 15276 pt (stock I think, overclocking is not mentioned) here: https://www.tweaktown.com/reviews/9839/intel-core-i7-11700k-cpu/index.html

you know that wiggling the mouse, or any kind of software running at the same time like temp monitoring software can introduce changes like that, and while 10% isn't exactly nothing, RAM speed differences, cooling differences, or background software can easily cause that

 

[lexluthermiester]

When package temp reached about 90 °C, I called it a day.

Oh, you're brave. I would call it good at 75C. I'm cautious like that. Don't like that much heat getting into my CPU's...(Yes, I know, that was an old post..)

 

[ThrashZone]

I don't want anything really but to understand why my Cinebench R23 stock score is that low compared to online benchmark values
I did what @ThrashZone suggested, setting Package Power Time Window to 448s instead of default 5-s, it doesn't change a thing, at best I got 14850 pt

Hi,
This thread is about all core boost clocks under advertised not why are my R23 scores lower than others on the net

To know that you have to run a cpu benchmark R23/... and watch hwinfo64 current clocks used during the entire benchmark and state what max clock were used and how many cores maintained the max clocks through the benchmark.

Lots of people lie on the net about what they do for the scores they get

Basic test,
Set all core to 4.6 or what ever your boost clock is higher if you want to ..5.0....
Enable XMP profile
Enable Multicore enhancement "remove all limits"
Leave core and cache voltage at auto.

Leave the timer on auto then change all four settings take notes of both runs.
Make sure to leave hwinfo64 on top so you can see what current clocks are doing through the test.

14850 means little not stating what the other score was.

 

[hymced]

Hi,
This thread is about all core boost clocks under advertised not why are my R23 scores lower than others on the net

To know that you have to run a cpu benchmark R23/... and watch hwinfo64 current clocks used during the entire benchmark and state what max clock were used and how many cores maintained the max clocks through the benchmark.

Lots of people lie on the net about what they do for the scores they get

Basic test,
Set all core to 4.6 or what ever your boost clock is higher if you want to ..5.0....
Enable XMP profile
Enable Multicore enhancement "remove all limits"
Leave core and cache voltage at auto.

Leave the timer on auto then change all four settings take notes of both runs.
Make sure to leave hwinfo64 on top so you can see what current clocks are doing through the test.

14850 means little not stating what the other score was.

I have run some more tests and I GET TO REACH A SOLID 15250 pt in R23 (max 15270 pt) on the 1st pass with AIO fan and case exhaust fan manually set to 100%, but WITHOUT XMP!

As soon as I enable XMP, R23 score drops to 14850 pt (max 14870 pt) (fans @ 100% again). R23 seems to be sensitive to temps... Now I don't get why memory affects R23 that much, everyone seems to agree it should not... Is it because of the higher memory voltage, and so greater temps of the memory controller? What do you think?

I think I don't saw it in my really first tests with XMP disabled by default because I wasn't careful enough to background processes eating some cycles.

In any case, all 8 cores are @ 100% 4600Mhz during R23 multicore

Now I am even more lost really...

 

[AusWolf]

I have run some more tests and I GET TO REACH A SOLID 15250 pt in R23 (max 15270 pt) on the 1st pass with AIO fan and case exhaust fan manually set to 100%, but WITHOUT XMP!

As soon as I enable XMP, R23 score drops to 14850 pt (max 14870 pt) (fans @ 100% again). R23 seems to be sensitive to temps... Now I don't get why memory affects R23 that much, everyone seems to agree it should not... Is it because of the higher memory voltage, and so greater temps of the memory controller? What do you think?

I think I don't saw it in my really first tests with XMP disabled by default because I wasn't careful enough to background processes eating some cycles.

In any case, all 8 cores are @ 100% 4600Mhz during R23 multicore

Now I am even more lost really...

It's weird that your score drops with XMP enabled. I'm not home this week, but I'll run some tests to see if that happens with my 11700 non-K when I'm back.

Though the real question is... does it matter?

 

[hymced]

I would say no it does matter at this point, but still it does no make any sense...

 

[CyberCT]

The best multi core I achieved with my 11700K in C23 was 14,786 with all cores limited to 4.6GHz. That was a month ago right before I swapped it for a 11900K.

Unless you have real good cooling, I'd recommend NOT setting the CPU core to a fixed ratio. Instead, in BIOS limit the max boost to 4.6 for each core (this way, the CPU will still lower clocks and voltage if there's little to no load). Then start messing with voltage offsets in the one BIOS screen. I could do -0.025 without any issues at the 4.6GHz point and the CPU consumed @ 1.2v in C23 in the multi core test. Passed all stress tests for hours. You will have to enable MCE and take out the other limits mentioned around this option in BIOS. Honestly, for gaming, the 11700K is decent enough.

My 11900K can do 4.8GHz at the same voltage as the 11700K's 4.6GHz for $100 more. I'd happily pay an additional $100 for 5.0GHz at @ 1.2v, since that's the max VCORE I'm comfortable with for this combo in the NZXT H1 and its 140mm AIO cooler. Max core temp was 75C. I build a new PC every 4-5 years so this will do for a while.

 

[ZenZimZaliben]

The best multi core I achieved with my 11700K in C23 was 14,786 with all cores limited to 4.6GHz. That was a month ago right before I swapped it for a 11900K.

Unless you have real good cooling, I'd recommend NOT setting the CPU core to a fixed ratio. Instead, in BIOS limit the max boost to 4.6 for each core (this way, the CPU will still lower clocks and voltage if there's little to no load). Then start messing with voltage offsets in the one BIOS screen. I could do -0.025 without any issues at the 4.6GHz point and the CPU consumed @ 1.2v in C23 in the multi core test. Passed all stress tests for hours. You will have to enable MCE and take out the other limits mentioned around this option in BIOS. Honestly, for gaming, the 11700K is decent enough.

My 11900K can do 4.8GHz at the same voltage as the 11700K's 4.6GHz for $100 more. I'd happily pay an additional $100 for 5.0GHz at @ 1.2v, since that's the max VCORE I'm comfortable with for this combo in the NZXT H1 and its 140mm AIO cooler. Max core temp was 75C. I build a new PC every 4-5 years so this will do for a while.

I'm running 5.2Ghz all cores and all the thermal management still works. If I'm on the desktop it's running at like 800Mhz. The 52x multiplier is just the max it will boost too but all cores run in tandem. Meaning it isn't a FIXED Ratio, it is allowed to speedstep, C1 all that thermal stuff. Just all cores clock exactly the same. In game, or intensive application all cores run at 5.2Ghz. Also 1.43V is on the extreme side for 5Ghz.