• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

DDR4 1.5v

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.


Nov 10, 2016
Well I updated my Bios to the newest Firmware 3007 and have been having Memory instabilities since. It's been four days and I've had 3 freezes and 1 bsod. Three of them were gaming and one was watching a you tube video. I had XMP off and had my Memory Overclocked just a tad to 3333 MHz. Everything Manual. Dram volatge at 1.36v After the last crash which was full on BSOD I turned XMP back on and ran my Memory at the 3200 MHz rated speed and have gone a day crash free but when I switched it back I accidentally switched the Dram voltage to auto after I turned XMP instead of before so instead of it automatically running at spec 1.35v it's been running at auto and my Motherboard has been feeding it 1.5v for a day. Just wondering if I should worry a little or if it's fine that it's been like that all day and night last night.
No worries in the least. People run these well past that. And its only been a day.

*waves hand* "This is not the problem you are looking for..."
Cool thanks. I'm gonna Flashback to 1902 or 2002 later tonight I think. No problems on 1902 for a couple months now until the Flash Sunday night. I was gonna throw a 7700k in here in a couple weeks is why but I gotta Friend who wants the Deluxe and 6700k So I might grab the m9F or check out the Asrock Supercarrier when it starts selling.

How do you like the M9F?
Cool. I forgot to mention I had a couple Q-code 22 halts after the Flash. I Flash Bios I lose my saved OC settings, set them all to where they were before.....code 22. Power off retry......success...running at 3333 MHz. A freeze...restart......code 22.....retry...success. Memory initialization. Downclock to 3200 with XMP boot first time been stable so far. One of them may even have been code 32 but definitely at least twice I've had code 22 since the flash.
I've heard on some boards the KL bios do not work as well with Skylake cpus. 22? You sure that isn't 55? That is a more common Memory code in my experiences...
Ya 22. I remember once I got up after a restart....sat back down spaced off for minute look over and it's just sitting there fans spinning code 22. I have seen that alot after reading around and hopping over to the ROG Fourm real quick that these KL compatible bios are causing problems with SL. So I figure I'll leave it at XMP 3200 for a bit and if nothing happens I'll test stability with KL. If I don't like what I see I'll sell it off and get a new board too. For no real reason but that I have a sickness and I want it. Lol :facepalm:
While I have your attention I wanted to ask one more question. Off topic. Nice review btw I just read it. There are a couple things that interest me about the Asrock SC, The 14 Power phases and the plx chip enabling two Gpu's in Sli to at x16 speeds. That is what I read today on their site anyway. Do you take a big hit running them at x8? I've never done Sli but may in the future for 4k. How big of a deal is it having that Plx chip and running two x16?
I have a very reliable source who says 1.8 is livable for DDR4. Well, he says DDR4 will survive it anyway. :)

Mmmm... will pump some more in mine, to see if I can lower the latencies :D

Edit: lowered to 13-16-16-35 instead of 14-16-16-35, with 1.55v... Notreally worth it for 0.07v, lol!
Last edited:
Johan45 told me that. He's running chilled water, and I don't know if that includes RAM. I doubt it, because RAM just doesn't get that hot
You can bench RAM at 1.8V but it's not recommended to run it 24/7 like that. About 1.5V will be max if you use it for longer and first HyperX Predator kits had even 1.50V XMP profile ( which was changed to 1.35V ). Memory will survive something higher but some users were reporting memory issues at high voltage. Johan is using these settings for benchmarks only. I had no issues with my Samsung/Hynix based kits up to ~1.85V but that was only for quick tests. Some Micron IC won't even boot when you set 1.55V+.
RAM voltage is separated from everything else. When you set high DIMM voltage then it won't affect any other component so it won't kill IMC as it has its own power source.

Re temps, here is screenshot from Crucial memory tests at 1.20V and 1.40V:
I had 1.5V+ too but can't find it now. Anyway you can count on 3-4*C more at 1.5V. Tests in link were performed on open rig without additional airflow.
Rainwater, just to make a point when flashing a BIOS. A previous stable overclock on one may no longer be stable on another.
Understood. In my Manual Code 55 is Memory not installed. 32-36 is Cpu Post Memory Initialization. I have had at least one code 32 (maybe two) and at least one code 22. I didn't find anything in my manual for 22 for some reason so I looked around online and I'm finding the same description. So these are the reasons I started with Memory. I slapped her with a quick 60 minutes of Aida 64 Extreme last night after enabling XMP. So far it seems that was the only OC setting that wasn't playing well. I don't see why it would make a difference if I run it at 16-18-18-36, 3200 MHz, 1.35v as spec manually or using XMP to achieve it (and maybe it doesn't) but I'm going to try to OC my memory a little more with XMP still enabled and see if it makes a difference.

I had it stable with Bios 1702 and 1902, but this 3007 doesn't seem to be stable with the same settings no. But the Q-Codes are why I am starting with Memory adjustments first. If instability continues I will explore deeper into it, or simply Flashback to 1902 or try 2202. Thank you...:)
Probably all X99/Z100/200 boards have AMI BIOS/UEFI with the same codes. 55 = memory can't be found or memory error. 20-30 errors generally mean IMC or something related like when you push memory too hard and it can't boot then you can see 23-3x ( I don't remember when they end as I usually see 23 ). Often you get this error when memory has too tight timings or one of the voltages is too low. Check VCCSA/IO voltages and/or VDIMM.
The easiest way to OC memory on Z170/270 is to enable XMP and set VCCSA ~1.3V, VCCIO ~1.25V, memory voltage at 1.35V and just go up with memory frequency. When it starts to freeze or you see errors then it will be about max stable without any special settings. Then you have a choice: bump voltages or lower memory ratio by 1 and tighten the timings.
You may ask why keep XMP enabled. It simply adjusts sub timings better on most memory kits.
Well I had a freeze last night around 2am. So I Flashed back to version 2202 immediately and went into bios set my Overclock up the exact same as it was......4.5ghz....1.3v.....LLC 7......Idle at 1.296v......jumps to 1.312-1.328v under max load.....Platinum running at 3333mhz for now.....booted into Windows 1st try and went straight to work.....Aida64E......just about to hit the 5 hour mark as we speak.......then on to some Memory overclocking or a heavier overclock on the Cpu....I am ordering a De-Lid tool this week.....perhaps even today.....I will look at some other threads for the best TIM to use and adhesive.....you are more than welcome to fill me in here LOL.....OK I just hit 5 hours......ending this test......saving settings in bios and upping back to 4.6 maybe even 4.7.....tho I do not like my temps at 4.7 at all....takes me about 1.42v for 4.7 Core/Uncore.....may drop uncore to 4.4 and see if I can get it under 1.4v load....Thanks....And no I can not explain why hw mon and real temp read my core temps at idle in the low 20's while my package reads 50c but at load package is close to core temps....:confused::shrug:
Screenshot (86).png