Z790 Chipset & Raptor Lake

[eliaden]

Amazing. Thank you for the guide.

onboard sounds works, copied the device properties for sound from valvator
didn't change the any kext

also valvator has a lot of unneeded stuff, so would start from my EFI, as it's cleaner and add to it if needed, but it's definitely a good source for anything else needed

Once I do further edits besides that I'll update my EFI post above

----

for cpu name, I needed to add RestrictEvents.kext (https://github.com/acidanthera/RestrictEvents) and set ProcessorType in OC config to 3842 3841
(thanks to valvator EFI post)

 

[dontblink]

onboard sounds works, copied the device properties for sound from valvator
didn't change the any kext

also valvator has a lot of unneeded stuff, so would start from my EFI, as it's cleaner and add to it if needed, but it's definitely a good source for anything else needed

Once I do further edits besides that I'll update my EFI post above

----

for cpu name, I needed to add RestrictEvents.kext (https://github.com/acidanthera/RestrictEvents) and set ProcessorType in OC config to 3842
(thanks to valvator EFI post)

Thanks again. if I may ask, what made you consider this board? Did you by any chance evaluate the aurous elite ax?

 

[eliaden]

Thanks again. if I may ask, what made you consider this board? Did you by any chance evaluate the aurous elite ax?

the extra USB C in the Back, and extra couple usb 3 in the back helped
and I think has additional nvme (not counting the PCI5 slot which I won't use)

the biggest thing was intel based ethernet port, as I read that Realtek and linux don't play well (learned that when selecting a micro PC to be an OPNSense router, maybe it was not linux but just OPNsense, oh well). And since this will likely become a linux machine in a few years, once macOS is "useless" on it wanted to play it safe

liked the look of it, but that's minor

 

[dontblink]

the extra USB C in the Back, and extra couple usb 3 in the back helped
and I think has additional nvme (not counting the PCI5 slot which I won't use)

the biggest thing was intel based ethernet port, as I read that Realtek and linux don't play well (learned that when selecting a micro PC to be an OPNSense router, maybe it was not linux but just OPNsense, oh well). And since this will likely become a linux machine in a few years, once macOS is "useless" on it wanted to play it safe

liked the look of it, but that's minor

Got it. Future proofing with the shift to Linux will be important for me as well. Thanks for the rationale.

 

[etorix]

the biggest thing was intel based ethernet port, as I read that Realtek and linux don't play well (learned that when selecting a micro PC to be an OPNSense router, maybe it was not linux but just OPNsense, oh well).

OPNsense is built on top of FreeBSD, not Linux.
But Realtek NICs do not play well with server duties, on any OS.

 

[dontblink]

the extra USB C in the Back, and extra couple usb 3 in the back helped
and I think has additional nvme (not counting the PCI5 slot which I won't use)

the biggest thing was intel based ethernet port, as I read that Realtek and linux don't play well (learned that when selecting a micro PC to be an OPNSense router, maybe it was not linux but just OPNsense, oh well). And since this will likely become a linux machine in a few years, once macOS is "useless" on it wanted to play it safe

liked the look of it, but that's minor

Another question for you.

Did you end up buying a CPU Frame? Curious what cooling solution you went with your rig given that from what I am reading on the internet it runs hot?

 

[eliaden]

Another question for you.

Did you end up buying a CPU Frame? Curious what cooling solution you went with your rig given that from what I am reading on the internet it runs hot?

CPU Frame ? I somehow didn't know those existed, but quick googling looks interesting, might look into it more and get one.

I have the noctua NH-D15, I eventually undervolted and underclocked the chip and it works well
I played it safe and set it to 5.2Ghz for all p-cores max boost and 4.2Ghz for the e-cores max boost. with -0.05v offset. All super stable, haven't had a random reboot even under high ram and high cpu usage (running a couple light linux VM while doing CAD and 3d rendering CAD project)

MacOS runs much cooler than windows, likely P-core not pushed as much, and pulls less watt (210w vs 240w in cinebench) (can barely hit 91C with cinebench or 3d rendering in fusion360, used to hit 100C instantly before the undervolt)

My single thread score did take a hit as I still need to set higher single core turbo speed and maybe do the same for 2,3,4 core at load at max speed, but it is by no means a sluggish system.

with the undervolt and lowered max top speed, in Windows hottest core is hitting 98~99C and no throttling with cinebench for 5min at least (and I assume linux would be similar), used to hit 100C in about 2seconds and throttle , so setting perfect for windows, if I pushed anything higher up It would throttle, but could get it to barely throttle and still be closer to 5.3Ghz and a slightly higher score, but decided to play it safe, with the same undervolt and better stability.

With current settings my current macOS cinebench R23 multi threaded score is ~36100 (which is about 1000 lower than before, but 10C lower and no throttling)

e-core would fail with that undervolt at 4.4Ghz, so 4.3 seemed too close but no crash, and settled at 4.2Ghz

The NH-D15 is clearly not good enough for the full potential of the chip, from my test it can handle about 240~250watt, if cores requested more it would simply hit 100C and throttle

But I have one core that runs much hotter than the rest (lowest cores is at 85C compared to hottest, and second hottest is usually 3~4C lower), so I might try to re-sit it.

but will play with it some more, after a re-sit and especially if getting a CPU Frame

 

[eliaden]

I bought the thermalright CPU Frame, so will give that a try soon

the test below are in windows, as it's easier to play with speed and the intel utility

I lowered the undervolt from -0.05 to -0.04 (tried -0.03 then back to -0.04 and so far no instability at -0.04, but if needed will go back down to -0.03)

Changed the max boost per core from 52 on all cores load count to 60x 60x 60x 60x 59x 59x 54x 52x (5 core load at 6Ghz would throttle to 5.9, so that's why I picked 5.9 on 5 core load)
kept the e-core max to 4.2ghz

Tried all e-core at 4.3 again, but they would force more v-core than needed on the p-core, and would end up throttling anyway to 4.2, and p-core would throttle to 5.1, so kept the 4.2.

Tried all core at 5.3 even up to 5.6, but after a few seconds will throttle to 5.2Ghz and after a minute oscillates between 5.2 and 5.1Ghz (due to the reduced undervolatge 5.2Ghz isn't as cool, but I get better single core performance, for a very small reduction of multicore performance, as I couldn't reach 6Ghz stable with -0.05v, but -0.04v seems stable in single core load)
This light throttling doesn't affect the e-core initially, but when p core throttle down to 5.1, e-core then went to 4.1Ghz
So nothing bad would happen at 5.6Ghz, and seems stable for the fraction of a second that it is running at that speed on all core load, but since it reduces to 5.2 anyway so quickly, just going with 5.2 for all core.

Running cinebench R23 and 2024 both in single core, keeps max core speed at 6.0Ghz and sometime down to 5.9 (just because more than 2 core are actually in use and if 5 cores are requested even if not all under load, would go down to 5.9 as the 5 core load setting)

Single run from a cold point gets about a score of 38500 then lowered to 37000 on 10min run.
Single run single core however is now excellent 2292 score in cinebench R23 (was getting 1995 before, when all core maxed out at 5.2 even in single core load) I don't expect it to lower on long runs as the temps are good and no throttling.

I need to make sure this is actually stable, but usually cinebench just gives an error or just closes when it's on the verge of stability, but not bad enough for a BSOD of worse instant reboot.

---

now back in macOS with those settings,

cinebench r23 getting multi core score single run is about 36000~36200, as expected since all core is still at 5.2 as before ( temps in macOS never reach 100C, still barely hit 90C)
Not sure why, but I don't think p-core are limited to e-core speed, and must be running at 5.2Ghz, but I don't understand the lower wattage compared to windows. in windows if running at 4.5Ghz I would get scores closer to 32000, so if p-core matched e-core at 4.2 I would be getting scores around 30000.
Single core is now back up to normal, at 2121 (was closer to 1950).

for cinbench 2024, multicore score hasn't changed as expected 2273, but single core increased from 117 to 130

 

[etorix]

P cores and E cores set their frequency independently of each other. What they share is the ring bus for exchanging data between them and with common caches, and this ring bus slows when E-cores are enabled.

Alder Lake – E-Cores, Ring Clock, and Hybrid Teething Troubles

This will be a short post about how Alder Lake’s ring behaves when E-Cores are active. With just P-Cores active, the ring runs at 4.7 GHz. But if anything is running on the E-Cores, the ring …

chipsandcheese.com

 

[eliaden]

Another question for you.

Did you end up buying a CPU Frame? Curious what cooling solution you went with your rig given that from what I am reading on the internet it runs hot?

OK so after removing the cpu cooler, there was a huge spot in the center with no thermal paste, I guess my CPU or cooler must have a light convex shape. (first time did a middle dot of thermal paste, this time I did a vertical line)

So that would have affected the temps for sure, so hard to tell how much the CPU frame helped, but I'm sure it did, but regardless for $10, totally worth it and such an easy install.

The CPU can now handle 5.4Ghz all core without throttling, lowest core at 75C and hottest at 92C
And could push the e-core to 4.3Ghz (4.4 would trigger throttling)
Wattage is now at 268watt

Then if I push all core load to 5.5Ghz, I'm getting throttling after 10seconds, so does allow a nice little 10second boost, but doesn't really drop, just a few blips down to 5.4, until after about a minute, then it settles around at 5.4Ghz, and pulling 300watt initially then throttling down to 290watt, so in that sense 290watt seems to be my new wattage limit with the noctua, which is a huge improvement over the 250watt limit I was hitting before.

And score in windows is shy of 40000 in cinebench R23 in single pass, 10min test got a 38500 score.

Also updated the max boost based on core load to 60 60 60 60 60 59 58 55
and kept the v-core offset to -0.04

then tried 62 62 62 62 60 59 58 55, but somehow doesn't stay steady at 6.2 on single core cinebench r23, and oscillates between 5.9 and 6.2, but doesn't show any throttling in the intel utility, but most of the time seems to stay around 6.0 to 6.1Ghz, but no crash nor any sign of instability. Cinebench R23 single score single run did increase and reached 2300 (windows).
So might try that for a little bit and see if I get any instability signs.

So turns out the noctua really isn't as bad as I thought, and would recommend it for an air cooler on the 14900K + thermalright cpu frame.
I kept using the noctua NT-H1 that came with the cooler for thermal paste.

Edit:
I don't know enough to know why, I'm guessing temp spikes were the cause before, but the CPU is now more stable with lowered vcore offset, I'm now doing some test at -0.07 (which would have instantly crash before I reapplied thermal paste+cpu frame), but no issue so far, and is stable at 5.5Ghz all core, pulling about 270watt at that vcore offset (in a cinbench R23 load) which the noctua can handle without throttling settling at 96C. (from further testing 285watt is where the noctua start struggling on my setup)
I now have 61 61 61 61 60 60 60 55 for turbo.
6.2Ghz was acting strange, so I gave up on that, but 6.1Ghz seems fine at least with vcore around -0.04 (but was limited to 5.4Ghz then on all core load), giving even higher single core score around 2375, so now seeing how stable it is with this new -0.07 vcore offset. But will settle at -0.06v to play safe since temps are still fine.

I get throttling for a fraction of a second as the fan ramp up at the start of a cpu load.