9900K or wait for KS?

shadow2761 said:
Nice. I will be installing it over the weekend on the Asus Strix Z390-E + G Skill 3600MHz C14 RAM.

I hope to hit 5GHz with temps around 80°C max with a H115i Platinum. I don't like the idea of being around 90°C.
Click to expand...

I'm using the same AIO for my 9900KF and it hits 80C when fully loaded running BOINC.

I run ring bus at 4.6ghz and AVX at 4.8ghz.

However, serious stress testing and it hits mid 90's C. I just play games on this rig so the H115i is plenty.
 
Dayaks said:
About two weeks ago I got a 9900KF. I didn’t see anything worth waiting for. I set it to 1.35V and 5Ghz all cores. I haven’t had time to screw with it further but seems completely stable.

For Timespy CPU tests I am at 70C with a custom loop set to 35C liquid temperature so I am happy with it. I’ve only gamed on it / encoded with Adobe Premiere Elements / Timespy.

Very happy with it. VR is much better on it than my 2700x.
Click to expand...

1.35v is too high. Most 9900's don't need that kind of voltage. Try lowering it to 1.25-6 and see how it does. Mine runs 5ghz at 1.26v.
 
DooKey said:
1.35v is too high. Most 9900's don't need that kind of voltage. Try lowering it to 1.25-6 and see how it does. Mine runs 5ghz at 1.26v.
Click to expand...

Will try. I think I just quickly glanced at a review. What is your favorite program to stress it?
 
Dayaks said:
Will try. I think I just quickly glanced at a review. What is your favorite program to stress it?
Click to expand...

I run the Intel stress test that's inside of XTU and I also run BOINC on all threads as well as try it with my favorite games. By doing all of those and it staying stable I consider that good enough.
 
my idea of stress testing is playing games for a few hours, because thats what my real world use will be. If it doesnt crash or behave abnornally I considered it passed.

I hate the idea of torture testing a CPU for hours at max temps just to see if its stable under those conditions.
 
shadow2761 said:
my idea of stress testing is playing games for a few hours, because thats what my real world use will be. If it doesnt crash or behave abnornally I considered it passed.

I hate the idea of torture testing a CPU for hours at max temps just to see if its stable under those conditions.
Click to expand...

The thing is though, it might be stable under some games and not others. If you can get it to pass some unrealistic torture test with a massive heat load, it will run like that under ANY real world condition.
 
Dan_D said:
The thing is though, it might be stable under some games and not others. If you can get it to pass some unrealistic torture test with a massive heat load, it will run like that under ANY real world condition.
Click to expand...

How do you feel about Aida64 + Prime95 as a good OC stresser?

Prime was always quick to call me out on incorrect voltage and Aida64 spotted memory errors quick-fast..
 
Dan_D said:
The thing is though, it might be stable under some games and not others. If you can get it to pass some unrealistic torture test with a massive heat load, it will run like that under ANY real world condition.
Click to expand...

Yes I understand if it is stable under the torture tests then it will be stable under any condition.

That's why I just play my games and if it is stable then I don't need to stuff around with ir, if it is not then I adjust the voltage/OC etc and continue gaming.

That way there is no need for me to put my poor chip under extreme heat for long periods.
 
shadow2761 said:
Yes I understand if it is stable under the torture tests then it will be stable under any condition.

That's why I just play my games and if it is stable then I don't need to stuff around with ir, if it is not then I adjust the voltage/OC etc and continue gaming.

That way there is no need for me to put my poor chip under extreme heat for long periods.
Click to expand...

You don't need excessively long stability testing. People will put systems under torture tests for 24 hours etc. but the fact is that's probably overkill. Anytime I've torture tested a system it's always crashed or locked up within the first hour or two. I can't recall ever seeing a machine run for 23 hours and then crapping out or anything like that.
 
DooKey said:
I run the Intel stress test that's inside of XTU and I also run BOINC on all threads as well as try it with my favorite games. By doing all of those and it staying stable I consider that good enough.
Click to expand...

I did the XTU test 2, 100% utilization at 5Ghz/4.7 uncore , 77C on the processor with 30C liquid. I went with 1.31V, 1.25V crashed under load. I’ll take it.

Pretty impressive it’s only around 140W on the processor. My 5960X took 400W to reach 4.8Ghz lol. Gotta love advancements.

So I am happy I went with the KF. Price seems right, performance is great, unless you want to wait a year (?) for 10-15% IPC (?).

I don’t see myself needing higher IPC or core count anytime soon.
 
Dayaks said:
I did the XTU test 2, 100% utilization at 5Ghz/4.7 uncore , 77C on the processor with 30C liquid. I went with 1.31V, 1.25V crashed under load. I’ll take it.

Pretty impressive it’s only around 140W on the processor. My 5960X took 400W to reach 4.8Ghz lol. Gotta love advancements.

So I am happy I went with the KF. Price seems right, performance is great, unless you want to wait a year (?) for 10-15% IPC (?).

I don’t see myself needing higher IPC or core count anytime soon.
Click to expand...

Good deal. At least you got that voltage down. I love my 9900KF for my gaming rig. Can't beat it.
 
I read I think on Gizmodo that when the KS hits the market the 9900K/KF/KS are going to all get a considerable price reduction. Do you think that's reliable info?
 
WildMonkey said:
I read I think on Gizmodo that when the KS hits the market the 9900K/KF/KS are going to all get a considerable price reduction. Do you think that's reliable info?
Click to expand...

Possibly. As it is the 9900KF can be had for as little as $419.99. I don't anticipate them dropping much below that, but I suppose its possible. Intel has done drastic price cuts in the past when its needed to.
 
I've had stability tests fail past the 10 hour mark. I do try for at least a 12 hour full test before putting a pc into service, I keep my machines for a few years so it's worth the extra few hours or day to ensure it's stable moving forward.
 
Keljian said:
200mhz is 5% performance. Tops.. just go the 9900k.
Click to expand...

There seems to be some weird scaling wall for most games past 5.0 ghz.
At 720p ultra, most games saw no performance improvement:
https://www.techpowerup.com/review/intel-core-i9-9900ks/13.html

1080p medium, which is probably more CPU intensive, showed little change as well.

Screenshot_20191111-232219_YouTube.jpg


130 fps is seems high, but remember, this is a 2080ti running 1080p medium.

The 9900k is geared towards hardcore gamers. The 9900ks is geared towards benchmark gurus and crazy people.
 
  • Like
Reactions: N4CR
like this
Nightfire said:
There seems to be some weird scaling wall for most games past 5.0 ghz.
At 720p ultra, most games saw no performance improvement:
https://www.techpowerup.com/review/intel-core-i9-9900ks/13.html

1080p medium, which is probably more CPU intensive, showed little change as well.

View attachment 199231

130 fps is seems high, but remember, this is a 2080ti running 1080p medium.

The 9900k is geared towards hardcore gamers. The 9900ks is geared towards benchmark gurus and crazy people.
Click to expand...

Interesting. Maybe something to do with u-arch? Maybe ring bus doesn't clock up with core frequency? I have no idea how it is linked internally or if it's separate like HT/IF etc.
So we could be seeing the limits of feeding those cores.. bit like Vega core overclocking being not 1:1 either.. memory limitations... this is similar behavior.
 
Nightfire said:
There seems to be some weird scaling wall for most games past 5.0 ghz.
At 720p ultra, most games saw no performance improvement:
https://www.techpowerup.com/review/intel-core-i9-9900ks/13.html

1080p medium, which is probably more CPU intensive, showed little change as well.

View attachment 199231

130 fps is seems high, but remember, this is a 2080ti running 1080p medium.

The 9900k is geared towards hardcore gamers. The 9900ks is geared towards benchmark gurus and crazy people.
Click to expand...


I'm not sure why they are running medium settings. That's leveraging the GPU more than running it on low. Honestly, that's likely why there is no real change past 5.0GHz. Next, 1080P isn't going to be more CPU intensive, if anything it will be slightly less so than a lower resolution. The more you increase the resolution and the more visual effects you apply the more your leveraging the GPU.
 
I currently have a 9900k and I like it but im also under a custom water loop and mainly game. The 9900ks is just a better bin for others that want that assurance kinda like what the did with the 8086k
 
My 9900KS is currently at 5.2GHz/1.30v - too high? I tried 1.275v and I got a BSOD.

Temps are great - maxing out at 74C (during Intel XTU) using an H115i AIO.

EDIT: lol up to 1.35v and prime95 still kills it. XTU passes. Raised VCCIN to 1.25.
EDIT2: looks like this CPU is already binned for 5GHz so beyond 5.1GHz you need more luck. I couldn't get true stability at 5.2GHz (went up to 1.35v). Just not worth it when I play online MMORPGs like ARK and it can royally fuck my game up lol.
 
Last edited:
Just for the record, 10% more processor performance at 5ghz is 500mhz, so 200 is like 4%or so.. is it really worth it? (4% faster processor is not going to mean the difference between playable and unplayable.. )
 
No it’s not pretty big. If it were 500mhz faster at stock that would be 10% as stated, it is 300mhz faster, which is just over 5%.
 
OK. I couldn't imagine anyone buying a 9900K today. It would make sense to get a 9900KS or go a step down to 9700K if it's a dollars issue. "Big" is relative when you're talking about top dog gaming performance. Even the 8700K could hang with the 9900K in most gaming workloads.

Beyond all of that - AMD is the best overall choice today. Further making the 9900K irrelevant.
 
Ok.

Here is my take:

the 9900k is about 25-35% more powerful in multithreaded apps than the 9700k, regardless of clock speed, and therefore if you’re using multithreaded apps it makes a lot more sense.

the 9900ks is (at best) 5% faster/more powerful than the 9900k therefore it is not significantly more powerful. What this means is that it is not as good value as the 9900k.
 
Last edited:
Can I plug a 9900K or 9900KS into a Z390 Asus ROG Maximus XI Hero motherboard, and it'll work okay? I think my delidded 8086K is starting to show problems, I used high voltage to get 5GHz. That or my day one 2080Ti is fizzling out, which is a real possibility.

I am stumped why my sig says Z370. Maybe the product on the vendor's page changed. What about Z370? I'll have to check what's in my PC, hah. First step is to downclock to 4.5 and see if the problem remains.
 
Last edited:
xorbe said:
Can I plug a 9900K or 9900KS into a Z390 Asus ROG Maximus XI Hero motherboard, and it'll work okay? I think my delidded 8086K is starting to show problems, I used high voltage to get 5GHz. That or my day one 2080Ti is fizzling out, which is a real possibility.

I am stumped why my sig says Z370. Maybe the product on the vendor's page changed. What about Z370? I'll have to check what's in my PC, hah. First step is to downclock to 4.5 and see if the problem remains.
Click to expand...

Yes K and KS will work fine in those boards. Your VRM's will be a bit toasty but it will work. Make sure you update to a recent BIOS first.
 
xorbe said:
Can I plug a 9900K or 9900KS into a Z390 Asus ROG Maximus XI Hero motherboard, and it'll work okay? I think my delidded 8086K is starting to show problems, I used high voltage to get 5GHz. That or my day one 2080Ti is fizzling out, which is a real possibility.

I am stumped why my sig says Z370. Maybe the product on the vendor's page changed. What about Z370? I'll have to check what's in my PC, hah. First step is to downclock to 4.5 and see if the problem remains.
Click to expand...

Probably because the there was a z370 version of the maximus hero. the X instead of the XI.

https://www.asus.com/au/Motherboards/ROG-MAXIMUS-X-HERO/

Suggest that you flash your bios and then upgrade.

As for VRMs getting hot, so long as you're not overclocking the snot out of it and have a little airflow, you'll be fine. There's about 400W worth of VRMs for a 240W peak processor.
 
  • Like
Reactions: xorbe
like this
The Microcenter in Dallas has a ton of 9900KS processors in the display case.
 
Dan_D said:
The Microcenter in Dallas has a ton of 9900KS processors in the display case.
Click to expand...

It says 6 in stock. I am guessing they will all be sold by year end (nationwide). 6 took a few days at my MN location.
 
My MC in Maryland is jam packed with Intel CPUs in general. Site shows 10+ 9900k/ks in stock each.
 
9900KS is in kind of confidential production, just to look good as to "we can make those things" to investors.
 
Jandor said:
9900KS is in kind of confidential production, just to look good as to "we can make those things" to investors.
Click to expand...

I don't think investors have much to do with it. The 9900KS is all about Intel trying to hold onto the one thing it has going for it and that's gaming performance.
 
You must log in or register to reply here.
Back
Top