• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

*** The Official Alder Lake owners thread ***

Vimes said:
@Kelt - have you checked your vcore voltage when it is running cinebench R23...? Not what you have dialled in, but actually running at when at full load overclock. The reason I ask is that when I had the MSI 690 Pro I needed to use adaptive voltage with a negative offset, or it would scale up quite high with just a manual vcore setting. I noted your Cinebench R23 screenshot in that other thread, and you are not far off thermal throttling, at 95c.

Just noting what you have written about entering 1.26v for your vcore.
Click to expand...

Set the vcore to 1.25V now, survived another 10 minute R23 multicore run, that's knocked a couple of degrees off! :D
 
@Kelt well done

When I had the MSI 690 Pro board, very impressive, I noted that the lowest vcore I could go to when running R23 would be fine and perfectly stable.
But I started to have some doubts when playing a game that used much less CPU overall load it would rarely exit to the desktop. So I raised the vcore slightly.
I noted that the vcore when under full load, and obviously at the 5.1 and 4Ghz, would then be at a lower vcore in a game that might be using the same Mhz speed but not full load.
I suppose that might be one factor as to why some people see great stability at full load but a potential problem at a much lesser load.
There can be other factors as well.
 
Vimes said:
@Kelt well done

When I had the MSI 690 Pro board, very impressive, I noted that the lowest vcore I could go to when running R23 would be fine and perfectly stable.
But I started to have some doubts when playing a game that used much less CPU overall load it would rarely exit to the desktop. So I raised the vcore slightly.
I noted that the vcore when under full load, and obviously at the 5.1 and 4Ghz, would then be at a lower vcore in a game that might be using the same Mhz speed but not full load.
I suppose that might be one factor as to why some people see great stability at full load but a potential problem at a much lesser load.
There can be other factors as well.
Click to expand...

Seems to be different on mine, I added the vcore to Afterburner's OSD, and it stays pretty much the same in games as when stressed ie around 1.252V.

Not had any CTD's yet.

5xxe0eE.jpg
 
Last edited:
Anyone know if the B660 or H670 chipset keep the K series CPUs running at full power (as long as it is cooled sufficiently) ?
I remember seeing a slide somewhere that non-K CPUs will reduce power after x time but I cannot remember if it is also linked to motherboard chipset.
 
Journey said:
Glad to hear they are being responsive, just be wary of they get you to send one only back they you want the same version returned.
Click to expand...


just got an RMA acceptance from Corsair. They include this in their instructions....

Your return merchandise authorization (RMA) has been approved, and we have included a return label as an attachment to this message. Please carefully follow the instructions below to ensure a smooth return. Note that if a Corsair agent has given different return instructions specific to your situation, those instructions should be followed instead.
All modules of memory kits being returned, even if not all are faulty. Due to compatibility concerns, we do NOT replace partial kits. Failure to return all modules will result in your replacement being delayed until we receive the missing modules
Click to expand...

seems good so far. Kinda impressed, rang them, they are based in America and the place of shipping the sticks to is the Bolton shop.
 
Vimes said:
just got an RMA acceptance from Corsair. They include this in their instructions....



seems good so far. Kinda impressed, rang them, they are based in America and the place of shipping the sticks to is the Bolton shop.
Click to expand...

Yeah that is their UK RMA centre, it is good news that it has been issued though and from my personal dealing I usually only takes 2-4 days, once they have them to you getting your new kit.
 
@Kelt you can see from your screen the very low wattage that it is drawing, very different from the 200w+ when running R23.
Is that with a fixed 1.25v vcore that does not change with the CPU speed and not an adaptive one which will vary..?
you did mention that you set a manual vcore.
 
@Kelt. Ah thought so.
I am trying to keep mine adaptive so that it varies from the 500Mhz to the max 5.1Ghz. Couldn't see the point of putting the max vcore into the CPU at its lower settings as it would be when at max.
As long as it is stable I'll be happy enough to use that.
 
I've just sent a 12900k back and grabbed a 12700k because the temps on the 12900k was really bad. I've gone from 70-90c during gaming / benching to a nice 40-60c on the 12700k. All in all im very happy with it.
 
Let's talk stability testing. Over the years I've gone through everything from weak (cpu z stress test) to insane (p95 small avx2). Long and short of it is that for ADL, I've settled on two main ones:

y-cruncher: https://www.softpedia.com/get/Science-CAD/y-cruncher.shtml

occt large/avx2 extreme: https://www.ocbase.com/

y-cruncher I use for an initial sanity test of cpu voltages, IMC voltages and mem timings. Launch the ycruncher.exe with a right click, run as admin
Then choose: benchmark (0), multi-threaded (1), and option 7 (2,500,000,000). If you fail, you're not stable. Simple as that. I run that about 4-5 times in a row. Have hwinfo64 open so you can make sure you're not thermal throttling as well.

occt large/avx2 extreme is used for longer testing of the entire system. This works buy hammering your ram, cpu, cache with rolling combinations of various instructions which induce large transients (the delta between high and low voltage). Because the test runs longer than YC (30mins for quick testing of changes and 1hr for something more daily focused), it tends to heat up components with sustained workloads which in itself can cause instability. This is also important as you want your PC stable in all scenarios where for a long game session or a sustained render and anything in between.

The above methodology has proven to be bullet proof for me with ADL and others I've recommended it to have had great results as well.
 
meh? said:
I've just sent a 12900k back and grabbed a 12700k because the temps on the 12900k was really bad. I've gone from 70-90c during gaming / benching to a nice 40-60c on the 12700k. All in all im very happy with it.
Click to expand...

Sounds like an thermal compound application issue - the 12900k does not run 30C cooler than the 12700k.
 
Kelt said:
All I've done so far is to set the core voltage manually to 1.26V, and the core multiplier to 51-P and 40-E.

Voila, rock solid 5.1/4 Ghz, with nice, cool and silent running.
Click to expand...

I've only tweaked with the voltage but when I do, it seems to underclock the cpu mhz slightly. Between 100-300mhz.
 
You must log in or register to reply here.
Back
Top Bottom