Intel Core i9-9900k 8c/16t, i7-9700K 8c/8t, i7-9600k 6c/6t 2nd Gen Coffee Lake CPU's + Z390

Well, at least it is nice to know the @Prema magic is working as expected. Never had a doubt, but confirmation is always good.

Did it run a bit on the warm side with the 8700K/8086K as well? With 2C/4T more, it would almost have to run hotter. (Just my redneck logic... not based on any science.)

 

He saw your concerns and enabled SVID so you could see the watts. I did not know he had SVID disabled and that, as we already know, will throw those numbers off every time. (Which is by design.) It did the same with my Maximus X Hero and 8700K as well.

 

Did you have any trouble getting Windows 7 installed? Saw a guy over at ROG forums pulling his hair out on that. I'd be pretty pissed off to find that Windoze OS X cancer was the only installable OS.

Here is what the watts show on my 7960X... with the fans turned off... Silent Heat... LOL.

 

Thank you for that update, much easier to read now and put into the context of the growing body of data about feeding power to the 9900k and the resulting thermals output for various motherboards.

If we look at the power output of only 163w at 100% load, it looks like there might be some power limits put on that configuration, either in design or the settings, which moderates the peak package / core temperature and keeps it below 90c's, 83c here @ 163w:

On the motherboards that will deliver 200w+ the full grandeur of the 9900k can be let loose, and then the temps continue to rise into the 100c range. As with this 360mm cooled gem running cores 107c/108c peak temperature @ 293w:

http://forum.notebookreview.com/thr...e-lake-cpus-z390.811225/page-85#post-10813565

This is why some motherboards like the "Asus Z390 Maximus Hero XI" are being used to show the tame side of the 9900k, the tame power throttled side that doesn't grow horns with high temperatures.
http://forum.notebookreview.com/thr...e-lake-cpus-z390.811225/page-77#post-10812082

Something to consider when attempting record breaking scores, find the motherboards that will allow full power operation, short term and long term. Probably with VRM under water cooling too.

Thanks for the update, and I am sure he will want to pursue unlocking the full power of the 9900k and then we can see the full power of the platform - and temperatures.

18:57 Performance wise does intel have a meaningful way to release a non HEDT chip before they transition to 10nm?

 

Still working out some things at the moment.

Haven't loaded up windows 7 yet.

 

At 4.8 temps rise pretty fast in the P870TM1

 

That looks like it's at 1.4V though for 4.8Ghz, or no? I'm guessing you don't need 1.4V for 4.8Ghz. And the Throttlestop pic is showing 4.7Ghz rather than the 4.8Ghz you mentioned?

 

That is the actual first test I ran. When I actually start trying to figure out the voltages, then I'll let you know what is optimal.
The picture also shows it throttling due to temps, but as you also see...no current limit throttling.


PS Under load voltage is 1.255V

 

Oh, ok, yes I thought that was a lot of unnecessary voltage! Will be good to see your final tweaked settings to see the optimum from your particular 9900K.

EDIT: Although with your "PS" Edit, 1.255V doesn't seem wildly over the top for 4.8Ghz, but no doubt you can tweak it down if that was just a stock first run.

 

i was juuuust starting to type my comment with exactly THIS when you ninja´d me

edit: man im slow today... keep us updated on your findings master john

 

Maybe this makes more sense.

 

5.090 Ghz
https://www.3dmark.com/3dm/29827413

 

im wondering if @Prema would also adjust the TjMax in his Bios mod to 115C to be able and more fully utilize higher 9900K clocks in the laptop form factor?

 

Sure, but the problem with doing this mod might bring on problems with other cpus....Speculations of course.

 

+rep. 9900K and Clevo

From @Talon testing (desktop)...


http://forum.notebookreview.com/thr...ers-welcome-too.810490/page-344#post-10813689

http://forum.notebookreview.com/thr...ers-welcome-too.810490/page-345#post-10813829

http://forum.notebookreview.com/thr...ers-welcome-too.810490/page-345#post-10813868

 

makes sense, this would have to be a 9900k exclusive mod then. or does the 115C also apply for the rest of the 9th series?

Sent from my Xiaomi Mi Max 2 (Oxygen) using Tapatalk

 

I surely do not have the answer on that just yet.

Auto Fans

 

I was wondering that too. Is Intel eating into tolerances to achieve higher clocks (to keep it g frequency competitive with 8700K/8086K) or is the process tweaked in some way (thicker silicon) to make it more resistant to temperature?

 

Or, has the ceiling actually been raised and we just didn't understand or realize it yet....

Speculations of course

 

interesting to see TS offering the option to switch TjMax to 115C if u uncheck PROCHOT, will it go back to the usual 100C?

 

Do you live in high ambient temp area/locality? Never seen 115C on a Intel CPU or are the sensors faulty?

 

Probably 9000 series processors with solder (STIM).

Read.... https://www.gigabyte.com/FileUpload/Global/multimedia/2/file/525/946.pdf

 

Yeah, that's a big 10-4 negative. You can not change that option using TS

It was just to show it was possible.

 

I find it interesting ASUS implementation of AI OC. It is a tend I think is, to a degree, positive for those that do not OC to get extra performance (although I don't see it beating manual settings any time soon).

EVGA implemented a step, stress and check in bios for x299 last year. Here, ASUS scores your cooling, and uses that score to recommend multipliers and voltages, learning from your usage, stress testing, etc., over time.

On AMD, precision boost overdrive acts as an advanced step down feature where it boosts to cooling limits up until something tells it no higher, then it goes down in smaller steps as more cores or heat is present, but better than Nvidia does on their throttle step own which can do a couple step down plateaus rather than their prior hard throttle downclock on pascal.

Personally, I don't like how tightly gripped Nvidia is on power delivery and voltages, but I do applaud the effort that went into their new OC tuning API.

I would prefer everyone learn and understand how to OC, what the settings are and mean, etc. But, allowing platforms to automatically evaluate silicon, cooling, etc., then adjust performance and settings as appropriate is interesting. Especially if they can program the iterative steps for optimization.

I also like the scale back feature where if cooling stops working as well, the frequency and voltage is pulled back.

Other than that, just sharing it as another video on the 9900K.

Sent from my SM-G900P using Tapatalk

 

Just in time for 9900K... Got this email today.

@Talon


The Ncore V1 and V1D finally available to pre-order!!!!


Hi

Sometime ago you wrote to us that you are interested in acquiring the Ncore. As we promised we are writing to you to inform you that, after months and months of preparation, multiple let downs from different design and website building companies we are happy to announce that:

OUR WEBSITE AND SHOP ARE "MOSTLY"
OPERATIONAL!

www.nudecnc.eu

If you would like to order you own NCore you finally can!!! I am so sorry that it took so long to get back to you all, I promise to make things go much smoother from now on. At the moment the final batches of Kickstarter Ncores are being finished, I hope to get first quarter of them next week. The Ncores purchased via the website will be hopefully delivered in December. Please remember that we are still working on our website improvments.

BUY NOW
NUDEcnc
Hampshire, SOUTHAMPTON
United Kingdom

You received this email because you signed up on our website or made a purchase from us.

 

Scary Mary Mother of Christ!

 

I remember reading about that over on Guru3D, I remember thinking that it was a cracking idea. I also remember reading in threads about the problems & delays, but I do want to see this come to market. And yep, 9900K - every bit counts when it comes to cooling!

 

Linus has a video on it as well. That is where I saw it first, but I was too late to pre-order. Probably good that I did now since I no longer have a use for it. I hope they will have one for HEDT processors, too. I would love to run my 7960X bare die.

 

Geez how underwhelming. Silicon Lottery bins are 4.9 to 5.1 (1.362V, -2 AVX). Not even 5.2.

https://siliconlottery.com/collections/coffeelake-r

 

give it some time, these first few offers are based on pretty small numbers. and dont forget: were talking about 8 core cpus now! u cant just scale up to more and more cores without it having any impact on clocks

the 5.1 Ghz bin is at 28%, more than high enough to leave room for a 5.2 Ghz bin, maybe around 7% of cpus judging by the current binning steps.

plus, they bin by using p95 so theyre veeeeery strict on stability. for most users thats sufficient to go up one or two bins for daily stability.

 

FYI...

10-28-2018, 06:28 PM Silicon Lottery Customer Support
Quote: Originally Posted by Sancus
" Were the 9700K bins thermally limited by the 240mm QVL or would the voltages required for 5.2 or 5.3 be too high/unstable even with 360mm AIOs or custom water cooling?"

"Yes, they were thermally limited. The 9900Ks are also going to be thermally limited. If temperatures can be kept down, they keep scaling with voltage like the 8th gen parts."

https://www.overclock.net/forum/180...fee-lake-refresh-binning-17.html#post27689750
https://siliconlottery.com/collections/all/intel-i7-9700k

 

Case King use maximum core voltage 1.35v as target. And 1.42v for Coffee lake (Prime95 26.6 with 1344K).

 

as SL stated in their forums, their binning process is not voltage but heat limited this time around. when heat is removed as a factor via supersized rads or chilling, then 9th series cpus scale just like the 8th gen

 

They still bin chips with over 1.35v vs. Case King. Not sure why Case King have put this target (in order to protect the silicon).

 

maybe there is no need. SL bins the cpus lidded and offers delidding service at 50 usd extra, whereas caseking sells all their bins delidded from the get go.

Sent from my Xiaomi Mi Max 2 (Oxygen) using Tapatalk

 

I’ll be home tomorrow afternoon and have about 5 days off to get back to testing. I had very limited time to get any testing done last weekend. Looking forwards to getting back to it and getting some testing in and @Papusan I know you wanted some stock will real long term voltage testing which I’ll get to.

 

So, this isn't to push AMD, rather to compare the performance of the 9900K to the new 12 and 24 core CPUs from AMD, just so people can weigh performance, etc. If putting in a laptop, obviously the TR series is not an alternative to the form factor desired. Instead, we can see better where the 9900K stacks up on performance vs true HEDT parts (including Intel's own 7820X and other HCC parts), since it is like a bridge between HEDT and mainstream. Also, there are other price considerations, such as MB, buying 4 DIMMs instead of two to get quad channel, etc.

But reiterating, this is only so that people can see where the 9000 series mainstream chips stack up against HEDT processors. Tomorrow, Hardware Unboxed will have the game review where the 9900K trounces the AMD CPUs, and I intend on posting that video as well (along, hopefully, with Steven's write up for TechSpot).



@ole!!! @Cass-Olé - one of you asked about rebooting going to local memory vs distributed. They added a dynamic feature for the 2920X or 2970WX where you do not have to reboot, but evidently the software is still buggy, LMAO. But, the feature is being worked on, so wanted to let you know since I was on gen 1 TR and wasn't for sure if it was changed or not.

Edit: here is OC3D write up on the CPUs. https://www.overclock3d.net/reviews...ration_threadripper_2920x_and_2970wx_review/2

 

Wow that 9900K looks balanced.

It’s great at pretty much everything. King in gaming and can hang just fine for productivity tasks. Once stock is plentiful and price drops back down like it did on the 8700K after launch it will make it even more appealing. Amazon had it available for preorder for $499 just a few days ago again. The $579 price is a gouge due to lack of stock and high demand.

 

It seems like retailers often do that just because they can. They know that the "gotta have it now" people are willing to step up and pay extra to be one of the first to have it. I think (hope) we will see the same with RTX once most of the 10-series overstock is sold out.

 

Except in the earnings call, Intel specifically said that the constraints on supply didn't really effect Q3, but is expected to effect Q4 more, meaning that December to Q1 2019 may be the time for prices to become more reasonable. This is what I pulled from the conference call (which I focused on long term rather than short term; but Q4 is still expected to have great revenue for Intel, so you may want to wait until after that to act on what is discussed here).

Reposted from a different forum, but written by me:


"This is what I pulled from the conference call, which all media covering Intel did not discuss:

1. constrained growth in Q4; lower end PC and IoT constrained the most
2. increased R&D to 13.5B or around there
3. constrained in 2019 because signals show that they can’t meet demand next year
4. Intel is looking to find what 14nm demand is to estimate level of constraint
5. depending on how progress on 7nm goes, they may change capex
6. logic capex up, memory down - so trying to get chips on 10nm and beyond
7. reiterated not too constrained PC side, Q4 is where constraints hit more
8. Cooper Lake still on for middle next year
   headwinds - growing competition, global trade (china and tariffs), wait and see, says manage and weather the dynamics blah

So, they clearly stated that the limits they have and constraints on supply start hitting this Q4, that they are worried about AMDs products into next year may dry up demand for 14nm designs (but that actually alleviates their constraints, as lower demand due to competition means they have better supply to meet that demand and therefore would be less constrained), that they are burning R&D to get 10nm working and are investigating if 7nm is working well enough to pull it in from risk production around 2021 with EUV also being examined for when they can deploy it, that they are focusing their expenditures getting working CPUs, not on memory products like Optane, and that next year may be a really bad year for them.

This came from statements in the Q&A session of the earnings call. ... https://seekingalpha.com/article/4214681-intel-intc-q3-2018-results-earnings-call-transcript

I ignore the fluff and go directly to what matters beyond next quarter. This is Intel, as nicely as they can, saying that they are fearful in a couple key regards. ...

But back on topic, if you match Intel’s statements with the rumor of 10nm being scrapped, they likely plan to call something else 10nm, but it will not be their originally planned process, they are looking at trying to jump ahead with how badly things are going on 10nm, and they are hoping their 14nm++ process will be competitive enough to hold AMD at bay, which will be using TSMC’s 7nm process.

....

Above, I posted from Intel’s 2017 press day documents showing that Intel, on their original timeline, predicted that 14nm++ would have a slight performance edge for transistors over 10nm+. Intel even said, at that time over a year and a half ago, that they were targeting 25% performance AND 50% power reduction for 10nm over their 14nm processes. We all know the chart with the power reduction at iso-performance or increased performance at isopower. Instead of giving the chart showing the line, Intel said their specific target on that graph.

Since then, we’ve seen Intel have to make changes to the 10nm process to try to get it working, along with countless setbacks. One account says Intel is focusing on the inclusion of materials like Cobalt and ruthenium for their process while others are focusing on the inclusion of EUV. Truth is, everyone will have to do both of those at some point to move to the 3nm and beyond designs. So nothing wrong with choosing one or the other, and Intel likely took the material sciences approach after being burned since 2015 on not having EUV ready to use.

So, focusing on the derision surrounding the “good progress” language missed out on what was said in the phone call and the likely state of affairs of the 10nm and 7nm processes at Intel and what may happen moving forward related to its process technology."

 

EVGA just listed a $999 RTX 2080 Ti yesterday. Sure the it’s lower binned “black” edition but it’s MSRP. It’s also out of stock lol. Still were getting there slowly. Hopefully after the holiday rush RTX prices will normalize a bit.

 

@ajc9988 thats good to see we can get a good feature being worked on.

@Mr. Fox you went to desktop just the right time, sl binning 9700k is around 5.1/5.2 and 9900k will be worse. the heat is crazy, this needed to be on 10nm or 14nm+++ but intel fked up

 

Well... here's my take on it...

It is easy to make assumptions and I suspect that the heat might have been just as bad or worse on a smaller die. I have a hard time believing that smaller will mean cooler, unless it's a weaker part that runs slower and uses less power. Most of us don't want that. I know I sure as hell do not.

We expect a lot in terms of clock speeds and that generates lots of heat. Hard to avoid that. Think of all the other "consumer" chips that cannot handle it, and if they could have, how hot and how much voltage might they have needed? AMD CPUs and most 6700K couldn't handle 5.0GHz+ gracefully either. I think we need to keep things in perspective, and I think the expectations for something magical with the soldered IHS were totally unrealistic. Remember, I questioned that all along and thought it was a bad idea. AMD soldered their best processors and they do not run cool when overclocked. Previously soldered Intel HEDT processors ran crazy hot when overclocked as well. We cannot keep doing the same thing and expect a different outcome. Or, can we? But, I'm glad the delid is possible without destroying the chip. Now it's just an inconvenient nuisance that they are soldered. Had it been harder or impossible to delid, then we'd all be mad about it.

What I am saying is, stop and think about it for a minute, and reflect on what we already know... 7700K and 8700K easily handled 5.0GHz with a nicely binned chip, but none of them ran cool without a delid. Add two more cores to what is essentially the same processor and guess what... gonna be way harder to cool. To me is just simple arithmetic and should be kind of expected. Anyone that was not expecting it to run hotter than 8700K was not paying attention and did not learn from recent history. 7700K was a little harder to cool than 6700K and 8700K yet a little more difficult than 7700K.

Plus, this is the 21st century. Hardly anyone does anything right. There is always much to be desired because we live in a day where being half-assed is status quo. Just look at all the turdbooks for the first clue. The second clue is how many people there are that think it's OK.

 

Double the cores vs. 7700K, 33% more cores vs 8700K, higher clocks and with somewhat usable temps. This with same tech as previous gen... I say, stop complaining. Intel could lower the clock speed (aka less heat) and still beat AMD's top dog Ryzen 2700X with a glance. Just buy a binned chips if you want most out of it.

 

this cpu is probably way too hot to run in p870tm unless we can a get a heatsink redesign, that internal water cooling.. hope its ready soon.

at best we'll run 9700k at like 4.8ghz, or 9900k at like 4.6ghz due to extra heat. honestly thats not much different from AMD's 8 cores at 4.2ghz especially with 7nm so close.

 

Yeah, I know bro. Totally agree with you. But, we're old guys and we tend to be pragmatic and use common sense. That's not how it works any more. The way it works today is you believe all of the unverified crap you read and hear, and then piss and moan when it turns out to be untrue. I think the complaining about 9900K running "too hot" doesn't reflect common sense. Otherwise, we should have been expecting worse and planned for.

 

The density issue is only going to get worse from here. This is why they are working on things like microchannels to cool 3D stacked chips, etc. But, that really is the beauty of TR, as it splits the cores up across a HUGE package area, that way heat isn't concentrated in one spot, which really can effect overall boost, among other things (also, you can bin the 8 core chips, then join the fastest two or four chips together for the final package; imagine Intel having two of the 8 core dies with a similar solution, but with that much more surface area for the IHS and the beefier cooling that comes with TR or even for Intel's X chips).

But I do agree. I did not understand why people expected miracles. I put the average clocks for the 9900K at 5.0-5.2GHz. Granted, the limitation isn't process but heat density, but the prediction nonetheless held up. It still is impressive engineering.

As to solder, I noticed a couple reviewers also seem to have gotten really good soldered CPUs, plus golden on voltages, as compared to some others. I was worried about the solder variance. But, so long as it can be fixed through delid....

@Papusan - the clocks are NOT higher for the average user. Granted it is heat and not voltage holding it back, and you get the extra performance from the two cores, but what is being seen is people having trouble at 5GHz on a number of chips on CLCs. If you ignore temp, then the higher clocks statement makes sense, but I caution the overstating of performance, as heat needs tamed first.

As to the 2700X statement, if you win by 20%, but charge 70% more, and street cost is 90% more, people will buy the mazda.

 

I've thought that all along. I've been telling myself " Self... nope, huh-uh, no way Jose" on that since it was first mentioned. Not gonna happen with the parts Clevo is using now. Or, maybe it will happen and it will just suck unless they make some major changes.