MSI Z490 Tomahawk + 10600k wont hit 4.8 at default

[connta]

Hello hello, all time reader, first time poster

I see there are a lot of threads about 10600k and overclock but I have a different problem so to not hijack anyones thread I will make my own.

The configuration is:

MSI Tomahawk Z490
10600k cooled by trusty old Megahalems C with Noctuas
Mushkin 3466Mhz 16-18-18 2x8GB

Everything is left on "Auto" except VDIMM which is 1.35, Enhanced Core Turbo which is Disabled and short/long power set to follow Intels spec as 125W long, 182W short, 56 sec boost time and 140A current. This to my understanding should be the "true" default by Intel spec.

First of all the processor wont hit 4.8 ever, 4.7 sure on 3-4 cores simultaneously but 4.8 never happens. I tried giving it some single thread loads but it looks like the newer windows scheduler brakes up even older games on multiple cores these days, which I clearly remember doing single thread only before some time. So I cannot reproduce true single thread workload and I'm thinking this could be the reason processor is not hitting 4.8. I've seen Gamers Nexus 10600k review and they say running Cinebench R20 single core mode should get you 4.8 on some cores (to verify boost is working correctly) but that is also not happening, dunno if they mess with affinity in task manager or what they do but its not working out for me. I managed to set 4.8 manually all core and it works but seems to need voltage over 1.25 which is a bit much for Megahalems as the temperatures become a bit toasty (but still bearable). Setting "Per core" overclock as 48/48/48/48/45/45 also does not get me 4.8 boost at any time, only if i set 49/49/48/47/45/45 then I get 4.8 and even 4.9 sometimes but that runs a bit unstable.

Now the questions, how would one go about testing if the processor max boost is working as it should? Also, I see everyone is doing all core overclock these days and no one is doing anything like per core 49/49/48/47/45/45 OC, why is that? Is this not possible/not recommended to follow Intels boost logic just to OC it a bit more than default?

And lastly, my voltages on Auto for VCCIO and VCCSA seem a bit excessive as 1.2 and 1.26 respectively. This is without any OC and from what I read those values are pretty high, even the motherboard itself labels anything above 1.15 and 1.2 as "red" when setting it manually so excess 0.05 on both. I tried lowering these to 1.14 and 1.19 and everything seems to be stable and temp also seems to drop a bit, what would you say these should be set on for default?

 

[bigtallanddopey]

Unfortunately. Technically it is 4.8Ghz max, the base frequency is the only guaranteed speed of any processor, although I am sure 100% will exceed the base frequency speed.

Will you be able to overclock past 4.8? I would imagine so, have you tried manual overclocking past that?

It could also be thermal throttling, the 4.8 is the max dependent on temps. So better cooling could mean you reach that number.

I have been seeing a few intel processors recently not hitting the max boost speed. Kind of like when Ryzen 3000 released and they couldn’t hit max turbo. But again they always say maximum or up to.

 

[connta]

Hey thanks for a fast replay.

I'm quite aware of the "up to" scam and it is how it is but what I'm asking is Cinebench R20 single core test truly the test where I should be seeing max default boost for certain (without any other modifications to processor affinity)? Because when I run that test and look at the cores activity in core temp or task manager it doesnt say that one core is doing all the work, it is in fact broken up on multiple cores doing some % of work so that makes me think that the processor then doesnt use "one core speed" but multiple core speed and does it at 4.7 which makes sense. If its by default only doing 4.8 on 1/2 cores loaded when you load all the cores it will revert a bit back. I'm just trying to get to the bottom of what exactly is happening.

Heres a screenshot https://ibb.co/hWZfZhL

Cinebench is run in single core mode an furthermore in advanced settings it is set to 1 thread only. The black line on task manager is where the test begins and as you can see not a single core actually goes back to a line as flat as idle, arguably cores 1, 2 and 4 do come close to a flat line (from top left to top right) but clearly theres core 7 and 8 doing majority of work but cores 9, 10, 11 and 12 are certainly not flatlining as on idle, so there is load on them (and no, its not background, as soon as the test is stopped all flatline).

Also, it is not the cooling, ye olde Megahalems is old but still very capable, in prime95 all cores loaded and AVX it runs at 4.5 and doesnt go over 80C. You can see core temp on the screenshot that temps are more than fine.

 

[bigtallanddopey]

I did see a video recently on YouTube but I cannot remember who it was or maybe more importantly whether it was an Intel or AMD cpu.

Anyway, what the video showed was the single core which was boosting, swapping as you described. I think the conclusion of this was that the cores swapped to basically spread the heat about. One core boosts, gets hot, as it gets hot the theoretical max frequency of that core then drops and so another core then becomes the “best” core and so that one starts boosting. It will do that until the single threaded work is complete.

My description may be bad but I think that describes exactly what you are experiencing. I shall try and dig out the video. It must be someone like gamers nexus or maybe hardware unboxed as they do those types of tests.

 

[connta]

Thing is, I'm not describing swapping of cores, that would look a lot different, like 100%, flatline, flatline, flatline... then flatline, 100%, flatline, flatline... then flatline, flatline, 100%, flatline...

Theres no such behavior, look at the screenshot, theres at least 2 cores doing 40% of the work each and the rest is 4 more cores doing like 5% each... thats not switching the load, thats distributing the load which seems to be exactly the issue here.

Anyway if you could find that video it would be appreciated, maybe it sheds some light.

 

[connta]

Found the answer in another thread https://www.overclockers.com/forums...-removal-and-VCore-Help?p=8148001#post8148001

Long story short, MSI is garbage. :/

 

[bigtallanddopey]

I couldn’t see the pic until now, yeh looks a little odd. I would however say to use something like HWmonitor to check out the cpu core usage etc. As core temp is only reporting the cores and not the threads unless you have disabled hyperthreading.

 

[connta]

Yeah, no.

Cant use either HWiNFO nor HWmonitor since the Nuvoton chip that handles that on this MB is very new and HWmonitor doesnt even support it so readings are not there at all and HWiNFO when used with this chip stops all fans dead, 0 RPM, lol.

All in all this is one of the worst motherboard experiences i ever had.

 

[bigtallanddopey]

Yeah, no.

Cant use either HWiNFO nor HWmonitor since the Nuvoton chip that handles that on this MB is very new and HWmonitor doesnt even support it so readings are not there at all and HWiNFO when used with this chip stops all fans dead, 0 RPM, lol.

All in all this is one of the worst motherboard experiences i ever had.

That is rubbish for you and in some ways very surprising. MSI have been producing some of the best boards, especially for AMD over the last few years. The B450 tomahawk was promoted and lauded by so many people. Maybe this is them not really supporting Intel at the moment, especially the Z490 platform which will be 1-2 years at best before it is replaced.

I did notice that hwmonitor hasn’t been updated since September last year, which is annoying as well.

 

[EarthDog]

I couldn’t see the pic until now, yeh looks a little odd. I would however say to use something like HWmonitor to check out the cpu core usage etc. As core temp is only reporting the cores and not the threads unless you have disabled hyperthreading.

threads don't have temps. They run off the cores.

I can't see the image either. OP, please host the pic at the forum instead of 3rd party.

 

[bigtallanddopey]

threads don't have temps. They run off the cores.

I can't see the image either. OP, please host the pic at the forum instead of 3rd party.

I was more referring to the workload. I was wondering if we could see if it was a core that was loaded or a thread from that core.

 

[connta]

I was more referring to the workload. I was wondering if we could see if it was a core that was loaded or a thread from that core.

It seems to be one core and one thread loaded for ~70% (35+35) of the load and the rest is scattered around... so not really a true single core workload still... dunno.

Ok, heres the picture, but basically it is solved now, it was just crappy MSI BIOS that doesnt like to be touched.

 

[trents]

"MSI is garbage." I don't think so. Be careful not to project your own anecdotal experience with one motherboard onto a whole company's line of products. As you can see from my Sig, I am running an MSI motherboard and I'm very pleased with it. Many others running MSI boards would agree with me.

 

[connta]

And the key in your statement is "I dont think so." Both are opinions and everyones entitled to one.

But I'll tell you why it is garbage.

Retail BIOS for motherboard came with only UEFI boot option. If you came from an older system with MBR boot you had to jump thru some hoops to even boot, install system on usb then convert your hdd MBR->GPT, good luck to Joe Average with that. Furthermore, the CMS compatible BIOS came only ~45 days later, after they probably got stormed with support request.

MSI Dragon Center is awful, on top of being a bad piece of software it was totally messing up my voltages when installed and crashed the Nuvoton control chip regularly so it would turn the fans off completely, only reboot helped. With a 3rd party program none of that happens.

BIOS that "messes up boost clocks when its touched too much" is not something anyone really wants to deal with, I personally had intermittent headache from it for days now until I came here and got an answer what is up with boost clocks. "It works, but you have to touch it just right" is not a great marketing slogan.

And did I mention that M.2 heatsink pads already ooze oil after bit more than a month of use? Well, they do. I wonder if the same pads are under primary heatsinks, hmmm...

If you dont think that is not enough to call a product "crap" then I dont know what is.

Its even more perplexing why would you attack my personal experience about a brand, if you like MSI use away, leave me to my disenchatment lol.

 

[EarthDog]

Nobody attacked you, relax. You can have your disenchantment.

I haven't run into any issues so far with MSI on Z490, X570 and B550. That doesn't mean they don't exist, but Woomack is on a whole different level of tweaking than most... so.... there's that.

 

[trents]

connta, don't get so defensive. I wasn't attacking you. I just wanted you to take a step back and look at the bigger picture before you trash a whole brand name. I dare say every board manufacturer has turned out a kinker on occasion. How many MSI motherboards have you owned anyway?

 

[connta]

Its all good you guys.

I just "stood my ground" and explained as to why I have such an opinion. I certainly wont consider MSI boards any longer, thats just a personal decision. They obviously lag in some parts, namely they have been critiqued for years (as i now find out) about not having OTA BIOS updates while all other manufacturers do this. Not to mention the reviewer arm twisting that brought out to light recently, it just erodes the brand if its coupled with lots of additional issues. I mean the list goes on and it might feel like bashing but it really isnt, its just how it is.

Once I have a bad personal experience with a brand I dont "give them a second chance", I change the brand, see what else is out there, sometimes it even turns out to be worse, just the way it goes, but you get better sense of the market going wide. In any case "to each their own", no harm in that.

 

[EarthDog]

Once I have a bad personal experience with a brand I dont "give them a second chance",

if i followed this, I couldn't buy anything, lol!

 

[connta]

You didnt get it man, no manufacturer is always great and none always suck. If MSI "burned" me this time, next time its gonna be Asus and after that someone else until we circle back to MSI.

Ofc if I would to blacklist a manufacturer forever after one fail there wouldnt be anything to buy, as you said.

BTW I hold MSI in the highest regards on GPU front, never had a problem with their GPU cards, au contraire. Currently I have a Gigabyte 1080ti that i got on the cheap during the mining craze but if there was a choice it would have been MSI without a doubt.

 

[EarthDog]

I totally got it. You clarified your point just above, however.

I get ya man. We all get upset over a bad apple or the way things can be from any board partner. I find it hilarious (from people who ACTUALLY do this) that people shun an entire brand because of a random bad experience.