Intel Core "Special Edition" Owners' Club

[Dr. Dro]

View attachment 365257

As it's usually the case, that's the second reason I went with the Apex Encore instead of the Tachyon X

 

[RootinTootinPootin]

View attachment 365257

same here for my Pro X..lmao..

 

[Makina]

Beta release looks like, gonna wait for the final one before updating.

 

[RootinTootinPootin]

Beta release looks like, gonna wait for the final one before updating.

nahh, I am still on an older BIOS, I don't really have issues and can't afford a performance loss joke on the new one, on the latest BIOS it forces C-States enabled on all configs, can't be disabled. I know, its not much of a deal breaker if they force it, typically no issues, just a slight penalty on latency since the cores would be idling now to C7 states (800mhz) and would take quite a few nanoseconds before it can go back up to full throttle.

 

[Macro Device]

There's a minor possibility I'm joining this club buying a 12900KS. The price is right but unsure if I can secure the quid for the deal.

I am but not this year apparently. Biz doesn't shine as bright as expected and I got things much more important than PC upgrades to take care of first. Yeah, running your own operation isn't as straightforward as being a salary man.

 

[Dr. Dro]

I am but not this year apparently. Biz doesn't shine as bright as expected and I got things much more important than PC upgrades to take care of first. Yeah, running your own operation isn't as straightforward as being a salary man.

Being your own boss is worth forgoing all PC upgrades in the world. Here's to hoping you catch a break my friend

 

[Macro Device]

It'll be fine. Been to much worse places than this. I just need to fix a couple things first.

 

[RootinTootinPootin]

new best

finally got 8600MT's dialed in, feels like I can do more with this P128 chip..

 

[Dr. Dro]

Nice! I got my hands on a liquid cooler again. The NH-D15S just isn't enough to tame the KS. This weekend i'll be putting some time to do maintenance on the PC. 7600 was so incredibly easy for me, that I'm confident that with the help of a CUDIMM I can probably get 10000 MT/s out of this board and CPU. Do you plan on getting a kit too? Just waiting for the Trident Z CK's to show up around here, hopefully early next year or mid-December (though that seems wishful thinking. The 48 GB 9600 MT/s kit should run 10K relatively easy.

Mind sharing your settings (timings) for the 8600 MT/s profile? I really wonder if my good ol' Hynix A kit can manage something like that, even if pushed to the limit, and it'd be a nice starting point for me.

 

[RootinTootinPootin]

hat I'm confident that with the help of a CUDIMM I can probably get 10000 MT/s out of this board and CPU

CU Dimm is a hit or miss, highly depends on BIOS, so far somebody tried it on Z790 and it works only on a specific bios (older) and it will also work only on "Bypass" mode, means it just works like an ordinary DDR5 DIMM, as to how far it went, it was just the same max frequency the tester can do with his chip, so unless Intel will allow a firmware update to let older platform users utilize the clock driver unit, its still advisable for Z790 and former platforms to use regular DDR5 Dimm

 

[Dr. Dro]

CU Dimm is a hit or miss, highly depends on BIOS, so far somebody tried it on Z790 and it works only on a specific bios (older) and it will also work only on "Bypass" mode, means it just works like an ordinary DDR5 DIMM.

Really? I thought it was supposed to be seamlessly backwards compatible (of course within limitations of the CPU IMC). Hope ASUS extends full support to at least the Apex Encore, we deserve that much

 

[RootinTootinPootin]

Really? I thought it was supposed to be seamlessly backwards compatible (of course within limitations of the CPU IMC). Hope ASUS extends full support to at least the Apex Encore, we deserve that much

backwards compatible YES, but needs more testing, and like I said on the 1st reply (edit) yields will be just the same, currently, CUDIMMS are on rarity and I saw on Amazon 1 G.Skill kit being sold for $1399 no way in hell I am paying that much for just a memory stick.

 

[Dr. Dro]

backwards compatible YES, but needs more testing, and like I said on the 1st reply (edit) yields will be just the same, currently, CUDIMMS are on rarity and I saw on Amazon 1 G.Skill kit being sold for $1399 no way in hell I am paying that much for just a memory stick.

Newegg has them on pre-sale at $389 for the high-end 9600 48GB kit, starts shipping Friday November 1

Are you a human?

www.newegg.com

It'll take some time to arrive here in the Brazilian market but hopefully, by then it's all tested and smoothed out. Much rather buy one of these since it'll also be used on my next build (unless it's DDR6 by then lol)

 

[Enigmatic9400]

Newegg has them on pre-sale at $389 for the high-end 9600 48GB kit, starts shipping Friday November 1

Are you a human?

www.newegg.com

It'll take some time to arrive here in the Brazilian market but hopefully, by then it's all tested and smoothed out. Much rather buy one of these since it'll also be used on my next build (unless it's DDR6 by then lol)

That...is quite expensive for a 48GB kit. Not sure if the speed is supported by your CPU.

 

[Dr. Dro]

That...is quite expensive for a 48GB kit. Not sure if the speed is supported by your CPU.

On a UDIMM, no CPU supports 9600 MT/s, that's why CUDIMMs are a thing. The highest performing UDIMMs are in 8000-8200 range, and that's where CUDIMMs start.

We'll really need these to become more available so people can test them more thoroughly on Raptor and Zen 5 platforms

 

[Enigmatic9400]

On a UDIMM, no CPU supports 9600 MT/s, that's why CUDIMMs are a thing. The highest performing UDIMMs are in 8000-8200 range, and that's where CUDIMMs start.

We'll really need these to become more available so people can test them more thoroughly on Raptor and Zen 5 platforms

Is CUDIMMS compatible with the Z790 platform for such high speed?

I do wonder about that.

 

[Dr. Dro]

Is CUDIMMS compatible with the Z790 platform for such high speed?

I do wonder about that.

CUDIMMs are supposed to be seamlessly compatible with UDIMMs, but like @RootinTootinPootin said there might be more to the story. If the clock driver ends up being useless on Z790 platforms (Apex Encore included), then I guess I should go after a decent 8200MT/s kit with no clock driver before they disappear.

My current RAM is fine for my needs, but I need another DDR5 kit for another build so might as well get something better soonish. I already started purchasing the parts for my next build

 

[RootinTootinPootin]

Is CUDIMMS compatible with the Z790 platform for such high speed?

High Speed RAM will always depend on the IMC Silicon Quality, from the person who tried the CUDIMM kit on the Z790 Apex Encore, it only booted on Bios ver 9905, as far as overclocking goes, it just did the same memory frequency his old kit did on the same chip, the kit ran on "Bypass mode", meaning, it never used the clock driver chip's capabilities.

If intel allows the use of CUDIMM on older platrforms such as the Z790, it will be a game changer. But I doubt they'll go that direction, ARL sales aren't looking good and currently they have been very silent on what's the next plan for the platform.

 

[Enigmatic9400]

Hi guys, I updated my Mobo to a Z790 Aorus Elite X Wifi 7 board.

Now the issue is, for the new 0x12B microcode update, Gigabyte basically disabled any all core overclocking potential by mandatorily limit all P core max turbo in adaptive voltage mode. I cannot boost anything over 56x in Cinebench R23 anymore even if I am not voltage, thermal, power, or current limited.

I tried disable intel default, undervolting AC LL and set 320W PL, the chip is pulling 295W in cinebench r23 while I set all p core turbo to 57x and turn TVB off, the chip will jump back to 56x as soon as cinebench r23 load start.

My question is:
1. Could I turn off this mandatory limit from gigabyte somewhere?
2. If I stay on the beta BIOS with 0x129 microcode along applying voltage limit of 1.45V and IccMax 400A with PL 320W, Will I still face degradations?

 

[Dr. Dro]

Hi guys, I updated my Mobo to a Z790 Aorus Elite X Wifi 7 board.

Now the issue is, for the new 0x12B microcode update, Gigabyte basically disabled any all core overclocking potential by mandatorily limit all P core max turbo in adaptive voltage mode. I cannot boost anything over 56x in Cinebench R23 anymore even if I am not voltage, thermal, power, or current limited.

I tried disable intel default, undervolting AC LL and set 320W PL, the chip is pulling 295W in cinebench r23 while I set all p core turbo to 57x and turn TVB off, the chip will jump back to 56x as soon as cinebench r23 load start.

My question is:
1. Could I turn off this mandatory limit from gigabyte somewhere?
2. If I stay on the beta BIOS with 0x129 microcode along applying voltage limit of 1.45V and IccMax 400A with PL 320W, Will I still face degradations?

1. This is probably motherboard dependent. Intel mandated motherboard makers to force enable C1E power state and remove the ability to toggle it with their latest BIOS wave, and they seem to have done so. I heard word of some additional restrictions being imposed by some motherboard makers, on my end, ASUS released an update for the Maximus boards which restores the ability to control this, I am not sure other boards have or will ever receive this feature back. Perhaps ThrottleStop can work around it, worth a try. I'm unsure if C1E being enabled is the source of your issue, but it is probably worth considering.

2. Could happen but I would argue it is unlikely, these are very similar to the settings I run daily (little less voltage) and I myself only flashed 0x12B BIOS 5 days ago when I reinstalled Windows during the New Year's break (truth be told, Genshin Impact was down for maintenance ), 0x129 was fine. The 0x129 update is what fixed the high voltage requests, so that was the bulk of the danger right there, 12B addressed the vmin shift problem. 12B is safer, but with care I do not think your CPU will die with 129.

 

[Enigmatic9400]

1. This is probably motherboard dependent. Intel mandated motherboard makers to force enable C1E power state and remove the ability to toggle it with their latest BIOS wave, and they seem to have done so. I heard word of some additional restrictions being imposed by some motherboard makers, on my end, ASUS released an update for the Maximus boards which restores the ability to control this, I am not sure other boards have or will ever receive this feature back. Perhaps ThrottleStop can work around it, worth a try. I'm unsure if C1E being enabled is the source of your issue, but it is probably worth considering.

2. Could happen but I would argue it is unlikely, these are very similar to the settings I run daily (little less voltage) and I myself only flashed 0x12B BIOS 5 days ago when I reinstalled Windows during the New Year's break (truth be told, Genshin Impact was down for maintenance ), 0x129 was fine. The 0x129 update is what fixed the high voltage requests, so that was the bulk of the danger right there, 12B addressed the vmin shift problem. 12B is safer, but with care I do not think your CPU will die with 129.

Thx for the reply.

Yeah I could try with disable c1e.

About the other point. My system almost never touches IA VR voltage limit 1.45V, even if I ran cinebench r15 for several runs consecutively, I did not see hwinfo64 has the limit triggered.

Also the new motherboard has less vdroop compare to my old one, which means I could undervolt below 300W to keep my chip stable in prime95 small fft.

I do see sometimes at idle, VR VOUT is around 1.35V while core frequency could drop to 3.4GHZ which maybe concerning for vmin shift.

Thx for the reply.

Yeah I could try with disable c1e.

About the other point. My system almost never touches IA VR voltage limit 1.45V, even if I ran cinebench r15 for several runs consecutively, I did not see hwinfo64 has the limit triggered.

Also the new motherboard has less vdroop compare to my old one, which means I could undervolt below 300W to keep my chip stable in prime95 small fft.

I do see sometimes at idle, VR VOUT is around 1.35V while core frequency could drop to 3.4GHZ which maybe concerning for vmin shift.

Now I am super confused, after flashing f8b directly from f7f bios, everything just works fine.

Maybe I just fixed a bug by flash the bios in the right order?