Soldato
Still not had the Alder lake Asus promotion cashback, 90 day limit is up at the end of the month for me.
-
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 ***
- Thread starter mrk
- Start date
Same. 99.9% sure we'll get it before the 90 days are up, as they'll not want thousands emailing the support address to investigate each case etc.
Same every generation, always small % difference in gaming between the flagship and the next SKU down. Though from coming up to 2 years experience with a CX48 4k OLED with first a 3080 and now a 3090, every frame matters when newer games can cause FPS dips.
IMO if you have a top tier GPU and can afford it, it's worth it. Using today's OCUK prices, its only £150 difference we're talking here between the i7 and i9. When the top tier GPU's cost £1050 for a 3080ti and £1400 for a 3090 flagship, it puts things into perspective.
Soldato
same ...
Soldato
I'll quote myself from the other thread.
Raptor Lake and Z790 will be out by the time you actually get paid.
Permabanned
- Joined
- 28 Sep 2018
- Posts
- 0
Weird I got my cashback for that and ratemygear already.
Associate
- Joined
- 17 May 2015
- Posts
- 20
Has anyone tried the LGA1700 washer mod?
Soldato
Yep, Made zero difference on my Asus board, From what I've seen it seems to be mainly Gigabyte boards affected as there is more than 1 company that makes the locking mechanism.
Welcome to the 12th club 
Ok so that will be the last BIOS update I ever do I think, if it ain't broke.....
The F7 release from the other day ended up being a dud. Appears it was causing me random BSODs, the error message each time was either page fault, kmode exception, dxkrnl.sys etc etc. The F7 BIOS changelog was posted earlier in the thread and it seems they changed the way the VRM is handled so must have been that I guess. I would see a Firefox tab just crash and it would want me to reload the tab, other software would crash and Windows reliability history was awash with programs that "stopped working".
Rolled back to version F6 and so far all back to normal.
The F7 release from the other day ended up being a dud. Appears it was causing me random BSODs, the error message each time was either page fault, kmode exception, dxkrnl.sys etc etc. The F7 BIOS changelog was posted earlier in the thread and it seems they changed the way the VRM is handled so must have been that I guess. I would see a Firefox tab just crash and it would want me to reload the tab, other software would crash and Windows reliability history was awash with programs that "stopped working".
Rolled back to version F6 and so far all back to normal.
Soldato
@mrk that is bizarre. I installed the same BIOS for my Gigabyte Aorus board, it seems "only" an incremental type BIOS change from the A and B variants previously, or so it seems. Point is tho, mentioned this before, I also suspect that they make changes that are sometimes undocumented.
As you changed BIOS revisions did you apply all your setting previously to this new sone..? Apart from loading my fan profile I manually re-entered all my previous settings from screenshots and all has been well.
BTW For my board they went from 6A to 6B and then this one to 6 as the final release, same change notes as yours, albeit they called your F7
As you changed BIOS revisions did you apply all your setting previously to this new sone..? Apart from loading my fan profile I manually re-entered all my previous settings from screenshots and all has been well.
BTW For my board they went from 6A to 6B and then this one to 6 as the final release, same change notes as yours, albeit they called your F7
Last edited:
I reloaded the fan profiles off USB but all settings were re-entered manually yep. It's got to the point that I know all the settings now off the top of my head so all is good there 
My usual process now is to flash new BIOS, it reboots 3 times as standard and automatically resets the BIOS settings, I then go into the BIOS and load optimised defaults, reboot and then back into the BIOS to enter previous settings manually.
I suspect you are right though, other changes were made to the F7 BIOS that were not listed in the changelog and it just so happens with my configuration (64GB 3600 in 2x32GB modules) etc it just wasn't a right match whereas on the F6 BIOS all seems well again. I have no issues with VRM or performance, so will just skip F7 I think and probably any new BIOS from here on unless there is a major change. Hopefully by then enough people will have fed back bugs to Gigabyte so the F7 issues causing my BSODs will be gone.
My usual process now is to flash new BIOS, it reboots 3 times as standard and automatically resets the BIOS settings, I then go into the BIOS and load optimised defaults, reboot and then back into the BIOS to enter previous settings manually.
I suspect you are right though, other changes were made to the F7 BIOS that were not listed in the changelog and it just so happens with my configuration (64GB 3600 in 2x32GB modules) etc it just wasn't a right match whereas on the F6 BIOS all seems well again. I have no issues with VRM or performance, so will just skip F7 I think and probably any new BIOS from here on unless there is a major change. Hopefully by then enough people will have fed back bugs to Gigabyte so the F7 issues causing my BSODs will be gone.
Soldato
Unless you have changed it, and I might have this wrong, didn't you at one point disable MCE within the BIOS as i seems to cause the CPU to heat up too much..? Do you still do that..?
I leave it on Auto, I only set it to enabled/disabled to see what it was all about and it instantly put the CPU to 100 degrees running Cinebench whereas it's in the 70s when set to Auto. The score difference was only 1000 on multi core so seems a bit pointless in the real world.
Soldato
You will probably find that it is boosting the CPU as needed BUT the issue isn't directly the boost but the auto vcore. Auto vcore voltages are just far too aggressive for most, if not all, motherboards that I have used. Of course doing anything different requires some manual intervention.
The same applies when you, and I have also, manually applied 1.3v to VCSSA voltage. If that is left on auto it applies way more.
The same applies when you, and I have also, manually applied 1.3v to VCSSA voltage. If that is left on auto it applies way more.
Auto for MCE definitely seems to be fine for me, gaming temps don't go above the 50s generally and when benchmarking it's in the low 60s. I don't know whether the F7 BIOS issue was related to MCE as both times it was kept on Auto, but I do know there were some fundamental changes in F7 as there were some new options I had not seen before like Initial Display Output being visible now which defaults to PCIE #1 so the monitor kicks in right away as opposed to previous BIOS versions where the monitor has a delay before it gets a signal on a cold boot up. The shutdown behaviour was also different, there's a longer delay after Windows has logged off and prepares to shutdown before the mobo actually powers down and the relay clocks.
Edit* Check out this whackness from F7 lol:
Edit* Check out this whackness from F7 lol:
Last edited:
Damn looks like my RAM isn't as stable as I thought, I did a new ramtest run and at 5300% I got one error. Would would be a recommended approach, increase DRAM voltage a bit more or system agent voltage? They are on 1.4v and 1.3v respectively still.
It seems at 3600MHz the board is the most unstable without tweaking voltages even more. The above test was done at 3400MHz as I wanted to test something out due to recent game crashes unrelated to undervolting the GPU as I tested at stock. RAM timings are at 18-22-22-42-64-2T.
Looking online, seems a lot of people have been having DDR4 stability quirks on these Gigabyte DDR4 Z690 boards and each person has had to tweak voltages differently. I appear to be one of the only people with 64GB in 2x32GB configuration on this board it seems.
It seems at 3600MHz the board is the most unstable without tweaking voltages even more. The above test was done at 3400MHz as I wanted to test something out due to recent game crashes unrelated to undervolting the GPU as I tested at stock. RAM timings are at 18-22-22-42-64-2T.
Looking online, seems a lot of people have been having DDR4 stability quirks on these Gigabyte DDR4 Z690 boards and each person has had to tweak voltages differently. I appear to be one of the only people with 64GB in 2x32GB configuration on this board it seems.
Last edited:
Soldato
Perhaps some of the issues you associated wit the F7 BIOS might have had the underlying RAM issues either made worse or not helping at all...?
Why don't you consider, for the benefit of testing, changing the Gear mode back to Auto or to 2, either way it will be at 2. The default, like mine, seems that Gigabyte have these boards we have set to 2. It will place less strain on the memory controller on the CPU. It might just help, for testing at least.
Mine is fine at Gear 1 but for 4000Mhz memory and only 16GB. But also my 12700k is overclocked. As mentioned the default is Gear 2 tho. That means, for me, my memory controller is running a 2000Mhz rather than 1000Mhz as it would with Gear 2.
BTW I have mentioned this before, with screenshots, that VCSSA on default puts more voltage through than setting it at 1.3v on my Gigabyte board. That is probably typical, like vcore on default.
Why don't you consider, for the benefit of testing, changing the Gear mode back to Auto or to 2, either way it will be at 2. The default, like mine, seems that Gigabyte have these boards we have set to 2. It will place less strain on the memory controller on the CPU. It might just help, for testing at least.
Mine is fine at Gear 1 but for 4000Mhz memory and only 16GB. But also my 12700k is overclocked. As mentioned the default is Gear 2 tho. That means, for me, my memory controller is running a 2000Mhz rather than 1000Mhz as it would with Gear 2.
BTW I have mentioned this before, with screenshots, that VCSSA on default puts more voltage through than setting it at 1.3v on my Gigabyte board. That is probably typical, like vcore on default.
It could be that F7 just amplified that underlying problem but lots of others seemed to have similar issues with F7 as well when I was looking around yesterday so there must be a fundamental change in F7 that brings about a few more quirks.
Hmm that is interesting yeah, the 1 error above in the karhu ram test was not far off the 99% coverage karhu states in their FAQ, now i cannot confirm if that 1 error was due to an actual memory error or a random bit flip due to something like cosmic rays (it does happen in this kind of setting and our RAM isn't ECC like server RAM is to account for these). I've increased DRAM voltage at 3400MHz for a new taste currently under way to 1.42v and will see if an error pops up at around the same coverage %. If it gets to 6400% without an error then I can consider that a fully stable configuration at 3400MHz at 1.42v and leaving SA at 1.3v. I can then keep that as a baseline and work my way up to 3600MHz slowly until the 2600MHz baseline is found.
I suspect this may be the only way.
Setting to gear mode 2 will half the memory bandwidth will it not? That's the main thing I was concerned about as this may give lower gaming performance or application performance in heavy workloads like Lightroom exporting and stuff whereas the difference between 3600MHz and 3400MHz is fairly small really.
If all else fails then I will drop to gear 2 and RAM at 3600 though I guess.
Hmm that is interesting yeah, the 1 error above in the karhu ram test was not far off the 99% coverage karhu states in their FAQ, now i cannot confirm if that 1 error was due to an actual memory error or a random bit flip due to something like cosmic rays (it does happen in this kind of setting and our RAM isn't ECC like server RAM is to account for these). I've increased DRAM voltage at 3400MHz for a new taste currently under way to 1.42v and will see if an error pops up at around the same coverage %. If it gets to 6400% without an error then I can consider that a fully stable configuration at 3400MHz at 1.42v and leaving SA at 1.3v. I can then keep that as a baseline and work my way up to 3600MHz slowly until the 2600MHz baseline is found.
I suspect this may be the only way.
Setting to gear mode 2 will half the memory bandwidth will it not? That's the main thing I was concerned about as this may give lower gaming performance or application performance in heavy workloads like Lightroom exporting and stuff whereas the difference between 3600MHz and 3400MHz is fairly small really.
If all else fails then I will drop to gear 2 and RAM at 3600 though I guess.