Newegg is in no position to risk another PR disaster so soon lol
Good to see Newegg can get it right sometimes! It is incredibly annoying when things are misadvertised like that though
I nearly fell in to the same trap with some Crucial Ballistix RAM because it was cheaper than the regular (but much more compatible) Crucial kits…
So you’re saying ballistix won’t work in frame.work?
It will work, but only at 2666MT/s, not 3200. It needs to be overclocked to 1.35V to achieve 3200MT/s, which the Framework (and actually most laptops) can’t (at least currently) do.
thanks for the heads up
No worries! ![]()
If it helps, I went for this… You can see the difference is that the voltage is 1.2V, and it’s 3200MT/s at CL22 which is a standard set of timings. I think 3200MT/s CL20 is also a standard, but I think it’s only Kingston Fury that is available and I can’t remember if I read of people having some issues with it not working quite right.
Hi, are there any plans to support XMP in the future or expose memory timings in the bios in the future?
Hi Aaron, those setting’s are locked by Intel. Framework has expressed no intention to try and circumvent those locks and I suspect any attempts to do so would likely get their allocation of CPU’s revoked.
I have 2x8GB Kingston Fury 3200 CL20 1.2v sticks and they work great. No issues. They benchmark great too.
I received my laptop and that specific RAM doesn’t work.
It runs if I put some compatible RAM in slot 0, but it runs at 2533 Mhz only then.
Didn’t think it would since HX429S17IBK2/32’s JEDEC/PnP is not at 3200MHz.
DDR4 @ 3200MHz is what’s supported by the processor (and the Framework laptop doesn’t use LPDDR4x):
Doesn’t matter at all.
It does none XMP 2933Mhz JDEC which any 11th Gen Intel CPU should handle just fine.
However it’s buggy and doesn’t. But still not to be expected.
It does if you want better performance. But sure, if that doesn’t matter to you, then it doesn’t matter to you.
I mean for the running at all part. ![]()
For now it’s fine so I can run it at least. Looking for new RAM atm and will sell the other one.
Hi guys, have anyone tried looking at RAM models that are 3200MHz without XMP and consumes less power?
From what I have been reading, Hynix sodimms consume less power than Micron ones.
Has anyone tested Timetec SODIMMs?
https://www.amazon.com/gp/product/B098TYN671
From what I have been reading, Hynix sodimms consume less power than Micron ones.
How could that be? I assume they both run at the JEDEC spec of 1.2 volts? The product you list certainly does
How could that be? I assume they both run at the JEDEC spec of 1.2 volts? The product you list certainly does
I read some posts online regarding real world observations by different brands of RAM such as this reddit post here.
Apparently, Micron dies kinda consumes more power compared to Hynix ones in general. Hynix also did previously boast about their new manufacturing process for DDR4 RAMs that consume less power, there are some RAMs that meet spec but fall on the lower range of the voltage tolerance.
There’s apparently some differences in idle power draw and load power draw due to the IC or something of the RAM. So I was wondering if anyone tested this further.
@Jieren_Zheng I followed some of those links and went down the reddit black hole for about 5 minutes. It’s clear that basically all of the posts are meaningless. As one commenter pointed out, going from one stick to two will obviously raise power draw as will increasing the amount of RAM (16 vs 32 GB) as there is more RAM to power. Another post related to the age of the RAM, I could conceivably see improvements in manufacturing processes lowering power draw but then Crucial would have improved their processes as well. There was no follow up to compare new micron to old micron. Overall, I think that the power draw difference, if any, would be minimal at best and you are far better off tuning your system in other ways, such as removing the many bits of software that Windows pre-installs that you may not want or ever need. Eliminating background processes and tuning CPU power draw by limiting clockspeed (if tolerable) would definitely yield results.
I don’t know of anyone testing this, I don’t mean to disparage the claim, it might be true for all I know. I just can’t think of any reason why it would be since all chips should follow the same standard.
Yea, I guess I will just go for the cheapest one that will fit then. I can’t find much data myself so I just wanted to ask around if anyone had some good experience.
Thanks.
It’s been a while, but is there any continuing work on this? I would like to be able to use my existing HyperX kit, but I’m currently limited by it not booting when Channel 1 (marked Channel 0 on the mainboard) is occupied by one of the HyperX SODIMMs