bscool wrote:
https://community.hwbot.org/topic/196740-rog-maximus-xii-apex/?do=findComment&comment=585112
Newer portable Memtweakit, have not tested much seems to display correct timings on z490 Apex using 11900k. Tried on z390 Hero and it works but displays 1t instead of 2t, may be other incorrect timings, but I didn't really look that close.
https://www.dropbox.com/s/xkwg0h0drlzy3aw/MemTweakIt.rar?dl=0&file_subpath=%2FMemTweakIt.exe
Hero XII with 10700k. Haven't thoroughly checked, but DRAM frequency. CR, tRAS, tRRD_sg and tRRD_dg, tcKE, tWCL are incorrect values, tWR is not reported. timings 3 and 4 seem alright,
I think this might work with 11th gen (or Z590). Not with 10th gen.
WISH IT WERE PROPERLY SUPPORTED.
F4-4000C19D-32GTZKK paired with a 10900K SP83 on bios 2004, which randomly or not, disappears from the Asus's support site or turns for a while into a final version and not beta as it was initially listed.
As expected in the Asus style, the voltages sometimes they go up, sometimes they go down, most often up, this time on the ascending slope compared to the previous bios, I lost 0.02 V on SA and IO for the 4000C16 profile.
F4-4000C19D-32GTZKK 16-16-16-36, VRAM 1.45V, VCCIO 1.26V and VCCSA 1.30V set in bios
TestMem5 3 cycles of tuned anta & GSAT for ~5hrs. I convinced myself once again that GSAT is the stuff to test DDR4 stability; usually when voltage is lost on the memory side due to bios, the SA voltage generally is the most affected, this time I had to raise the IO as well, saw this with GSAT only.
Out of boredom and lack of occupation I did some "tuning", enough not to lose any voltage on RAM or MC, I reduced the tRFC to more than half (had it 700), possible under 300 but I probably I'll have to readjust the SA to be fully stable, tREFI I lowered it to be as safe as possible and to keep bandwidth and latency still ok, tFAW just for fun 4xtRRDS, tWR 14 - and the rest of the secondary and tertiary latencies left on auto since I saw no visible changes modifying them. As a side note, tXP 4, PPD 0.
Well... 2x16 on Rocket Lake seems to be not so fun anymore, latencies with Rocket going through the roof, those Gears are meh, so it looks that I'm forced to stay with the crappy Comet for a while, 11900K looks a bit dodgy so far, and it doesn't seem to be day & night in games compared to 10900K, some users say that 11990K would be even weaker, noticing a regression, which is either due to latency between cores, and with the microcode update to be more or less on the same level with 10900K the voltage gets a bit high... for now Rocket hasn't convinced me, although there are some gamers who say that lows would be a bit better than Comet, but I would take this with some salt, whatever...
I'm thinking to switch just the mobo with a Z590, or better to keep the Z490 with bios updates only, without ME firmware: D, I've learned the hard way not to update ME as soon as it appears, there are too few MEs for performance, when I had the Z390 I saw this clearly, there were reactions on the forums as well, I'm keeping a copy or two of bios, who knows when the Asus page will be down again, saw they tend to withdrawn even some bios versions.
Alder Lake is likely to be a similar milestone as the one that was from Z97 to Z170, with that Skylake odyssey, DDR4, nvme & ****, forced to switch on W10 , beta bioses for months plus wasted time and to make the stuff working =))
As expected in the Asus style, the voltages sometimes they go up, sometimes they go down, most often up, this time on the ascending slope compared to the previous bios, I lost 0.02 V on SA and IO for the 4000C16 profile.
F4-4000C19D-32GTZKK 16-16-16-36, VRAM 1.45V, VCCIO 1.26V and VCCSA 1.30V set in bios
TestMem5 3 cycles of tuned anta & GSAT for ~5hrs. I convinced myself once again that GSAT is the stuff to test DDR4 stability; usually when voltage is lost on the memory side due to bios, the SA voltage generally is the most affected, this time I had to raise the IO as well, saw this with GSAT only.
Out of boredom and lack of occupation I did some "tuning", enough not to lose any voltage on RAM or MC, I reduced the tRFC to more than half (had it 700), possible under 300 but I probably I'll have to readjust the SA to be fully stable, tREFI I lowered it to be as safe as possible and to keep bandwidth and latency still ok, tFAW just for fun 4xtRRDS, tWR 14 - and the rest of the secondary and tertiary latencies left on auto since I saw no visible changes modifying them. As a side note, tXP 4, PPD 0.
Well... 2x16 on Rocket Lake seems to be not so fun anymore, latencies with Rocket going through the roof, those Gears are meh, so it looks that I'm forced to stay with the crappy Comet for a while, 11900K looks a bit dodgy so far, and it doesn't seem to be day & night in games compared to 10900K, some users say that 11990K would be even weaker, noticing a regression, which is either due to latency between cores, and with the microcode update to be more or less on the same level with 10900K the voltage gets a bit high... for now Rocket hasn't convinced me, although there are some gamers who say that lows would be a bit better than Comet, but I would take this with some salt, whatever...
I'm thinking to switch just the mobo with a Z590, or better to keep the Z490 with bios updates only, without ME firmware: D, I've learned the hard way not to update ME as soon as it appears, there are too few MEs for performance, when I had the Z390 I saw this clearly, there were reactions on the forums as well, I'm keeping a copy or two of bios, who knows when the Asus page will be down again, saw they tend to withdrawn even some bios versions.
Alder Lake is likely to be a similar milestone as the one that was from Z97 to Z170, with that Skylake odyssey, DDR4, nvme & ****, forced to switch on W10 , beta bioses for months plus wasted time and to make the stuff working =))
