Thank you! I'm getting great single core results so far, but my multithreaded results are a little low compared what I've seen on other benchmark sites. E.g. geekbench 5 I am running ~2000 points lower than i should be.That's a very nice build and upgrade.
There appears to be too many blank boxes even for the free aida64. I could be wrong...This is not my rig. Taken from a post on OC.net. Memory latency w/DDR4 4266 @ cl14 SR is around 40ns. Saved a pic of it,
View attachment 52900
That is probably ~the limit of Gear1 for SR. Above RKL in clock, but also way above 10th gen in latency.This is not my rig. Taken from a post on OC.net. Memory latency w/DDR4 4266 @ cl14 SR is around 40ns. Saved a pic of it,
Nah, when tuning memory, one runs tools multiple times, and Aida takes minutes for full run.Photoshop ???
Whoa, thanks! Didn't actually know that but I wish I had when I tuned my memroyNah, when tuning memory, one runs tools multiple times, and Aida takes minutes for full run.
So it has a shortcut - You can double click either exact cell or row title and only those will be calculated.
Nice rig! Very clean!I've just finished my new build - first gaming PC in a while. It was a hard toss up between this or AMD, but I felt that the CPU gain of the i5 vs the 5600/5800 was too material to overlook, especially for something I plan on keeping for a couple of years with the possibility to upgrade to DDR5 once prices become more reasonable.
A huge upgrade to the Dell XPS 15 I've been running in clamshell for a while (i7 9750h). So far I am very impressed.
- i5-12600k
- Asus Prime WiFi board
- 16GB Corsair DDR4 RAM (Don't see the point spending too much on DDR5 currently)
- Crucial P5 1TB M.2 (Again, don't see the need for a PCIE 4 drive just yet)
- 6700XT
- Corsair H100i Elite CAPELLIX cooler
View attachment 52867
My ddr5-4800 is not stable at 5800DDR5 5200 looks doable with my 32GB 4800 Micron memory. Going to do some longer term testing.
Have you tried playing with the uncore for a little bit more performance. You're running very high frequencies that could benefit from high uncore as well. I'm sure it'll take some voltage and that's what you're preventing. May also impact your tuning.5200 was not totally stable at 1.25V. Don't want to push voltage too high on these DIMMs. 5000 with primary timings of 38-38-38-80 are stable.
Good enough until DDR5 6000+ materializes.
I also backed down a speed bin on the 12900K for 24/7 use. Even though 5.3GHz on two cores and 5.2GHz on all cores is stable in all my torture tests, I just like the piece of mind knowing 5.2GHz on two cores and 5.1GHz on all cores is going to run everything I throw at it. It will be using less power as well.
The ECores are at 4GHz two cores and 3.9GHz all core. Uncore/L3 at stock 3.6GHz.
I just wish my new house was ready so I could hook it up to my home theater setup for its full intended purpose, gaming. 65" LG OLED with Klipsch floor standing Atmos surround speakers.
Build was smooth once I updated the BIOS. My motherboard came with the initial BIOS.
View attachment 53284
I built the system last Friday and everything went smoothly. Then all of a sudden on Sunday it started getting very unstable. It stopped recognizing the boot drive and if it would get into Windows it would immediately freeze, restart, or return some BSD. So I'm thinking BIOS update, bad boot drive, bad motherboard, or some weird software conflict from a Windows update. Turns out after a few hours of troubleshooting it was the BIOS.Yes, my Asus board BIOS also needed updating! Lovely set up
I think I went overkill with my cooler vs yours and you went i7 vs my i5...
Where did you learn that from? Officially supported RAM is validated on the platform, they use the actual products and check stability at target frequency and timings provided by the memory manufacturer. The XMP profiles are stored on the memory sticks, and DDR4 supports 2 profiles. The memory manufacturer can set two different profiles in there, one of them being less aggressive to maximize compatibility with different hardware.I also learned something. XMP I profile on my memory was tuned by ASUS and compared to the XMP II profile the ASUS engineers just relaxed a few settings. So I guess when RAM is officially supported it means the engineers have created a profile they feel is stable with that board. Am I on the right track here?
Memory vendors test their modules following Intel’s Intel XMP test plans, in addition to their own. Test results are recorded along with the specific processor, motherboard, and BIOS version used. Once passing test logs are reviewed by Intel modules are considered for addition to the Intel XMP certified list.
In the BIOS on my ASUS board when you hover over the XMP settings the info pop up that appears at the bottom of the screen says that XMP I are the memory timings tuned by ASUS. I don't remember the exact verbiage but I can check it if you want. Then I looked at the timings of XMP I and XMP II in the BIOS and noticed they are very close but some XMP I settings are slightly relaxed. I am thinking that XMP I is validated by the manufacturer but not XMP II. In my case the 4 primary timings and RAM speed are the same but there are small differences in other timings.Where did you learn that from? Officially supported RAM is validated on the platform, they use the actual products and check stability at target frequency and timings provided by the memory manufacturer. The XMP profiles are stored on the memory sticks, and DDR4 supports 2 profiles. The memory manufacturer can set two different profiles in there, one of them being less aggressive to maximize compatibility with different hardware.
DDR5 has up to 5 XMP profiles, 3 can be set by the memory vendor and 2 by the user. More details here, including information about Intel's XMP certification program: