Soldato
If the 2700X could do 4.2GHz all core at 1.00v then it would have a lower TDP. It can't so AMD gave it a higher TDP to allow it to hit the higher boost clocks.
Re-read what was said this is not about cost of running.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
I9 9900k
- Thread starter Nelly
- Start date
If the 2700X could do 4.2GHz all core at 1.00v then it would have a lower TDP. It can't so AMD gave it a higher TDP to allow it to hit the higher boost clocks.
Re-read what was said this is not about cost of running.
Associate
Well board and cpu just arrived, looks like the processor was produced in week 35, so between 27th Aug - 2nd Sept, and number 80 on the wafer, some used to say this was an indicator of how well the processor would OC, not so sure myself, we will see, lol.
So if Intel can match AMD at a lower TDP and on larger lithography without better efficiency, how did they do it? Did they add some magic unicorn dust to their STIM?
Intel's 14nm++ is probably more efficient than GF's 12nm. Remember that Intel's headline node name is already more conservative than most other foundries and Coffee Lake Refresh is the 5th time they've used 14nm (Broadwell -> Skylake -> Kaby Lake -> Coffee Lake -> Coffee Lake Refresh). Although the names are similar, 14nm++ is a huge refinement on 14nm.
Remember that only place where Intel was/is truthful is their process tech. 14nm is 14nm, even 10nm was 10nm, before it died under the weight of cobalt. TSMC, Samsung and GF use different metrics to name their processes and then they add a shovel of PR and we have 14nm, 16nm, 7nm, etc.
On paper Intels 10nm was very similar if not better in characteristics than TSMCs 7nm, because Intel was always very aggressive when developing and refining process tech. Now imagine you have all this experience working with single process for so many years now, you end up with miraculous 14nm++++++. Though at some point this won't be enough, well currently process already feels old in a way. I think 9900K is the pinnacle of their 14nm development, and next iteration if that arrives will see minimum gains, or even negative gains. 14nm can't compete with 7nm from foundries.
Look at the heat output of 9900K, its insane for Desktop CPUs, that's in a way a biproduct of sticking to 14nm for too long.
On paper Intels 10nm was very similar if not better in characteristics than TSMCs 7nm, because Intel was always very aggressive when developing and refining process tech. Now imagine you have all this experience working with single process for so many years now, you end up with miraculous 14nm++++++. Though at some point this won't be enough, well currently process already feels old in a way. I think 9900K is the pinnacle of their 14nm development, and next iteration if that arrives will see minimum gains, or even negative gains. 14nm can't compete with 7nm from foundries.
Look at the heat output of 9900K, its insane for Desktop CPUs, that's in a way a biproduct of sticking to 14nm for too long.
Logically speaking, and aside from the fact that no one talks about the cost of electricity when using the word 'efficient' on it's own, the actual cost of electricity is completely arbitrary and could be rendered negligible or vitally important on a political whim. Power efficiency is not arbitrary, and recognised universally to the same unwavering degree. It doesn't get much more mathematically sound.
Soldato
GF isn't going to do 7nm. And TSMC is already preparing for 7nm EUV in 2020 (Zen 3 and Apple). That's mile ahead than the 10nm.
LOL you people are highly amusing
All i said was the hardware unboxed guy has some kind of agenda... i didn't attack him like it was claimed or question or argue or disagree with any of his data.. in fact i made NO comment at all about the content in his videos all i said is he has an agenda with some of his 9900k videos
what that agenda is who knows maybe he is trying to create controversy to get more views
maybe he is trying to gain favour with AMD fans and Ryzen owners so they sub to his channel who knows and when i really think about it i dont really care anyway
and searching my previous posts to see if i bought a 9900k or find out if i posted anything about the 95w TDP? seriously?
go ahead search away ill be interested to see what you find
You took the bait from the AMD fanboys. Better off enjoying what you buy without having to justify it to the budget and power conscious. As for the video, it was clearly agenda driven.
7nm EUV is technically 7nm+ or ++ depending on if they can pull it off so soon. I was comparing initial 7nm vs 10nm from intel.
We are forgetting when Intel was promising us 10nm
If everything was going according to plan, today we would be at far end of 10nm+ easily.I wasn't even mentioning GF, when I said "foundries" I meant everyone except Intel
Since 10nm from Intel is dead, by the time they bring 10nm+ (whatever they gonna name it), which might be more relaxed specd process, hopefully AMD will have the balls and finances to be among the first 7nm EUV customers with their designs, to keep pressure on Intel
Associate
- Joined
- 30 Aug 2018
- Posts
- 2,483
What settings are people using for their 9900k overclocks?
I have seen (on reddit
) everything from 5.3 @1.25 (throttled so not actually 5.3) and 5GHz at 1.2ish to 4.9GHz at over 1.35.
Sometimes this is with an AVX offset, and in some cases it isn't.
Der8auer suggests a -4 offset for AVX in his overclocking video.
If those of you with the 9900k could post your own settings/ experiences so i know what sort of ballpark frequency and voltage i should be going for (as well as avx offset) i would really appreciate it.
I have seen (on reddit
) everything from 5.3 @1.25 (throttled so not actually 5.3) and 5GHz at 1.2ish to 4.9GHz at over 1.35.Sometimes this is with an AVX offset, and in some cases it isn't.
Der8auer suggests a -4 offset for AVX in his overclocking video.
If those of you with the 9900k could post your own settings/ experiences so i know what sort of ballpark frequency and voltage i should be going for (as well as avx offset) i would really appreciate it.
Soldato
What settings are people using for their 9900k overclocks?
I have seen (on reddit
Sometimes this is with an AVX offset, and in some cases it isn't.
Der8auer suggests a -4 offset for AVX in his overclocking video.
If those of you with the 9900k could post your own settings/ experiences so i know what sort of ballpark frequency and voltage i should be going for (as well as avx offset) i would really appreciate it.
Cinebench R15 stable with 4.7ghz cache and 4133mhz C12-11-11-28-1t memory
water temp of 24c
LLC6
set voltage is 0.04v lower than below actual voltages(max read voltages with HWmonitor)
5ghz 1.29v
5.1ghz 1.34v
5.2ghz 1.39v
5.3ghz 1.44v
running 24/7 rock solid 51/45/3600 1.35v max temp in game(BF4) with water approx 25c is around 59c
Associate
- Joined
- 30 Aug 2018
- Posts
- 2,483
Thank you!
Does this pass prime with or without avx and do you have an offset for that?
I know prime is an unrealistic workload.
I've just run cinebench 4.9 1.25 and then 5GHz crashed so I jumped to 1.3 for that. But I'm using a -3 offset right now for avx so running prime the cores drop to 4.7.
I haven't tested any lower voltages with 5GHz or gone higher than 5 yet because my PC is still in bits while i sort the loop out.
Edit - the voltages I'm listing for my OC are the ones i set in Bios. I have Hwinfo open but I'm not 100% sure where to look to check what voltage is being reported.
Soldato
i run my 24/7 settings with a -2 AVX offset and yes it passes prime 26.6 and any other synthetic. the benchmark settings i run 0 offset as theres no AVX!!
max XTU here http://hwbot.org/submission/3974322_nickolp1974_xtu_core_i9_9900k_3574_marks
max XTU here http://hwbot.org/submission/3974322_nickolp1974_xtu_core_i9_9900k_3574_marks
Soldato
Ramtest / HCI stable at those settings?
You should aim for stability at 4000/4133 with the Gene, it can handle those speeds just fine. CAS16 should be possible with reasonable voltage (with the sticks you have).
Last edited:
Associate
- Joined
- 30 Aug 2018
- Posts
- 2,483
This info is really helpful, thank you very much!
Associate
- Joined
- 30 Aug 2018
- Posts
- 2,483
How is your rig?Ramtest / HCI stable at those settings?
Is it still at 5GHz 1.25? I cant remember if you had an avx offset or not?
Soldato
What at 4133 12-11-11-28-1t you know it's not!!Ramtest / HCI stable at those settings?
You should aim for stability at 4000/4133 with the Gene, it can handle those speeds just fine. CAS16 should be possible with reasonable voltage (with the sticks you have).
5ghz all core, no avx offset. Adaptive voltage loading at 1.296v GPU RTX 2080ti shares loop.
Hottest core:
Destiny 2: 59 (average 45-55) 1440p
1hr OCCT large: 62
1hr Prime95 avx: 85 (Coolest 79, may try reseat?)
All rads of the thicker variety. 360x2 + 240.
Haven't tried higher than 5ghz
Hottest core:
Destiny 2: 59 (average 45-55) 1440p
1hr OCCT large: 62
1hr Prime95 avx: 85 (Coolest 79, may try reseat?)
All rads of the thicker variety. 360x2 + 240.
Haven't tried higher than 5ghz
Associate
- Joined
- 30 Aug 2018
- Posts
- 2,483
Thank you!
Soldato
Hi, yes, around 1.25v at load. AVX Offset is 1
lol, no, but you could aim higher at 4133. CAS 16/17 with that board potentially