• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Lexar NM790 (4TB) made my PC go back to Windows XP days, since it caused my PC to be SO slow and laggy!

Joined
Feb 15, 2009
Messages
705 (0.12/day)
Location
Norway, which means Amazon is not available...
System Name Winter v3.2024
Processor Intel i7 12700K (since november 2021)
Motherboard Gigabyte Z690 Gaming X (since november 2021)
Cooling Air Liquid Freezer II 360 with LGA1700 kit (since november 2021)
Memory Crucial Ballistix 2x16gb 3600mhz C16 (since november 2021)
Video Card(s) Gigabyte RTX 3060 Ti Gaming OC Pro LHR - Rev3.0 (since july 2022)
Storage 1x Kingston NV2 2 TB m.2 nVMe (since december 2024), 1x Seagate 4TB SATA
Display(s) 1x 55" LG C1 4k OLED, 1x Gigabyte 32" M32Q and 2x AOC 27" CG1
Case Fractal Design Define R6 (since 2018 and still working like a charm!)
Power Supply Corsair RM850x black (since august 2022)
Mouse Razor Deathadder v2 (since december 2021)
Keyboard Varmilo VEA109 v2 MX Silent Red (since august 2022)
Software Windows 11 Pro
*topic*

I plugged in the Lexar M.2 and now, my PC is super slow and laggy. Just opening notepad took me almost 20 secs! :O

What am I missing here? Apparently, I have the newest firmware for the NM790, as shown in Lexar DiskMaster. Ohh, I am running my Windows 11 on my M.2 1TB drive (see my specs). I have also not allocated the Lexar M.2, because my PC is too slow... It was working great before I added the NM790.

I have not touched BIOS yet. Maybe something there, or?

How do I fix this problemo, o' supreme overlords?

Here's what it says in Lexar DiskMaster:
1278361782367.jpg


PS: Apologies if this was posted in the wrong section! I did not notice it until now... :/
 
If you haven’t even allocated it shouldn’t affect system performance. So might be some PCI-e shenanigans going on with incorrect lane assignments. But overall the whole issue of “systems feels poor after adding an SSD” is extremely vague.
 
Well, I tried to allocate and format, but no success.

Hooowever! I might have found the evil culprit... Not sure how this happened, but apparently, this fella has come loose! Does it mean that my GPU is the one causing this by not being correctly in place, or?

IMG_20250721_203641.jpg

EDIT: After I cleaned the dust (that you can see traces of in the photo) on them connectors, plugged it in thoroughly, it's now working like a charm again! PHEW!
EDIT 2: Nevermind, it's back at it again......... :( What to dooo? :(
 
Last edited:
use powersettingsexplorer - make sure your nvme powerstates ms is 15 or below. Likely the powerstates are cuasing the lag.
 
use powersettingsexplorer - make sure your nvme powerstates ms is 15 or below. Likely the powerstates are cuasing the lag.
Ok, but where do i locate the powerstate for my nvme in that app? ...

Also, the m.2 didnt come with anything that i should remove on the stick itself, right? Cuz i did not remove/peel off anything on it...
 
*topic*

I plugged in the Lexar M.2 and now, my PC is super slow and laggy. Just opening notepad took me almost 20 secs! :O

What am I missing here? Apparently, I have the newest firmware for the NM790, as shown in Lexar DiskMaster. Ohh, I am running my Windows 11 on my M.2 1TB drive (see my specs). I have also not allocated the Lexar M.2, because my PC is too slow... It was working great before I added the NM790.

I have not touched BIOS yet. Maybe something there, or?

How do I fix this problemo, o' supreme overlords?

Here's what it says in Lexar DiskMaster:
View attachment 408696


PS: Apologies if this was posted in the wrong section! I did not notice it until now... :/
This is odd. Even with this drive being DRAMless, it should still perform better than you seem to be experiencing. What happens when you put your old drive back in?
 
This is odd. Even with this drive being DRAMless, it should still perform better than you seem to be experiencing. What happens when you put your old drive back in?
I did not remove any drives. I have one M.2 already there, which is where my windows 11 is. I just added the NM790 to the middle port in those 3 available ports (under the big silver heat protector), and that is when all hell broke loose....

I did not notice any sticker on the Lexar, so should i take out and see if it had one and then remove it? Hmmm...
 
That clip is supposed to stop what happened to you from happening, namely PCIe bus issues caused by lanes making bad contact, no contact, or even cross-wiring between lanes.

If your GPU sits firmly in the slot and you never move or bump your PC then it's probably fine - but if that clip is snapped or broken, remove the clip from the other PCIe x16 slot you're not using and put that on the one with your GPU in it.
 
That clip is supposed to stop what happened to you from happening, namely PCIe bus issues caused by lanes making bad contact, no contact, or even cross-wiring between lanes.

If your GPU sits firmly in the slot and you never move or bump your PC then it's probably fine - but if that clip is snapped or broken, remove the clip from the other PCIe x16 slot you're not using and put that on the one with your GPU in it.
Ok, I will try this.

But as you can see, it will be difficult to fit it to the upper slot without smashing these cables... What are these attached cables anyway? Maybe I should remove those cables and then attach them again when I have fiddled the card into place, or? ... Hmm.... IMG_20250721_232612.jpg
 
I did not remove any drives. I have one M.2 already there, which is where my windows 11 is. I just added the NM790 to the middle port in those 3 available ports (under the big silver heat protector), and that is when all hell broke loose....

I did not notice any sticker on the Lexar, so should i take out and see if it had one and then remove it? Hmmm...
Ok, fair enough. So what happens when you remove the drive from your system? Do things improve?
 
I hate to be the annoying one, but if all the other suggestions fail:

Update mobo BIOS (if you know what you are doing)
Update firmware on the nvme drive itself
backup everything offline in cold storage
unplug cold storage
clean install windows
install chipset drivers
reboot
install gpu drivers
reboot
let windows do all its updates
reboot
and if that doesn't fix it, welp i don't know what to say
 
Ok, fair enough. So what happens when you remove the drive from your system? Do things improve?

I hate to be the annoying one, but if all the other suggestions fail:

Update mobo BIOS (if you know what you are doing)
Update firmware on the nvme drive itself
backup everything offline in cold storage
unplug cold storage
clean install windows
install chipset drivers
reboot
install gpu drivers
reboot
let windows do all its updates
reboot
and if that doesn't fix it, welp i don't know what to say
This noob earned some new EXP: So, the BLUE STICKER on the thermal cover (that is supposed to "attach" itself onto the M.2 stick) was ON, because I'd never opened that thermal cover before, and I completely forgot to remove the sticker before attaching the M.2 drive!

Secondly, I was able to put the "lock" for the GPU slot back in place, as you can see and hear in this video, so now the GPU is not being abnormal and weird on me, seemingly anyways... I checked if the loose piece had something missing, but only the bottom was a bit "scraped off", but not something that would be meaningful in terms of attaching it in place again: https://streamable.com/ammk1p


For now, it seems like everything's solved, but if not, I will proceed to the steps that Space Lynx mentioned.
 
This noob earned some new EXP: So, the BLUE STICKER on the thermal cover (that is supposed to "attach" itself onto the M.2 stick) was ON, because I'd never opened that thermal cover before, and I completely forgot to remove the sticker before attaching the M.2 drive!

Secondly, I was able to put the "lock" for the GPU slot back in place, as you can see and hear in this video, so now the GPU is not being abnormal and weird on me, seemingly anyways... I checked if the loose piece had something missing, but only the bottom was a bit "scraped off", but not something that would be meaningful in terms of attaching it in place again: https://streamable.com/ammk1p


For now, it seems like everything's solved, but if not, I will proceed to the steps that Space Lynx mentioned.
That blue sticker is just a plastic film to keep dust off the sticky thermal pad until it's ready to be used. It wouldn't cause the slowdown weirdness, it would just limit the duration the drive could run burst transfers for before it started to thermal throttle.

Your fixed GPU slot is the solution. I refurbish workstations monthly and there *is* such a thing as a bad case. If the relationship between the motherboard tray and the rear expansion slot panel is anything other than perfectly aligned, I'll need to flex the GPU in the slot slightly to get the screws in for the graphics card. This is bad for the motherboard, bad for the graphics card, and over the next 3-5 years of thermal cycling can lead to failures exactly like the ones you described, or worse - I've seen cracked PCBs!

Even it all seems nice and solid when I'm building it up I've just come to accept that minor misalignments between the motherboard slot and the case slot are more trouble than they're worth in the long term, so I scrap the case and buy a new one - there's no point jeapordising the future stability of a €1500 workstation for the sake of reusing an old ATX case worth next to nothing.
 
I wrote several times to not buy NVME without DRAM below those NVME tests.

No DRAM
Quote:
  • No DRAM cache (but still performs extremely well)
  • Low random write and mixed IO performance
  • Some thermal throttling when heavily loaded



A proper Drive is e.g. a Kingston KC3000. I read it in a forum ages ago and therefore bought it.

In my point of view: Your lexar drive is just a backup drive. For backups where wirte and read speeds do not matter.

--

How to fix?

You may look into how to teak the host memory buffer option.

You use the SYSTEM RAM for basic drive operations = host memory buffer.
 
Last edited:
Your fixed GPU slot is the solution. I refurbish workstations monthly and there *is* such a thing as a bad case. If the relationship between the motherboard tray and the rear expansion slot panel is anything other than perfectly aligned, I'll need to flex the GPU in the slot slightly to get the screws in for the graphics card. This is bad for the motherboard, bad for the graphics card, and over the next 3-5 years of thermal cycling can lead to failures exactly like the ones you described, or worse - I've seen cracked PCBs!
Yup. It absolutely is the case that a poor PCI contact can cause weird lane issues and unexpected behavior from the system. I know everyone just got used to PCs being LEGOs and stuff, but the physical connection in the interface is still that - physical. It must be well seated to avoid issues, especially these days since, perhaps contraintuitively for some, the more advanced and speedy the bus the more sensitive it is to poor connectivity.

I wrote several times to not buy NVME without DRAM below those NVME tests.
It barely matters these days with modern controllers, no, and the tests bear this out. The drive itself is absolutely fine and isn’t OPs issue. He experienced his slowdown when the drive wasn’t even allocated and no data from it was or even could be accessed by the OS. Or even existed, for that matter.
 
This noob earned some new EXP: So, the BLUE STICKER on the thermal cover (that is supposed to "attach" itself onto the M.2 stick) was ON, because I'd never opened that thermal cover before, and I completely forgot to remove the sticker before attaching the M.2 drive!

Secondly, I was able to put the "lock" for the GPU slot back in place, as you can see and hear in this video, so now the GPU is not being abnormal and weird on me, seemingly anyways... I checked if the loose piece had something missing, but only the bottom was a bit "scraped off", but not something that would be meaningful in terms of attaching it in place again: https://streamable.com/ammk1p


For now, it seems like everything's solved, but if not, I will proceed to the steps that Space Lynx mentioned.
You live and you learn. We've all been there at some point. Glad you got things sorted & working!
 
Last edited:
I wrote several times to not buy NVME without DRAM below those NVME tests.

No DRAM
Quote:
  • No DRAM cache (but still performs extremely well)
  • Low random write and mixed IO performance
  • Some thermal throttling when heavily loaded



A proper Drive is e.g. a Kingston KC3000. I read it in a forum ages ago and therefore bought it.

In my point of view: Your lexar drive is just a backup drive. For backups where wirte and read speeds do not matter.

--

How to fix?

You may look into how to teak the host memory buffer option.

You use the SYSTEM RAM for basic drive operations = host memory buffer.
What you say is true, but DRAMless drives really are fine for Windows and Linux. So many drives on the market are DRAMless and it's less about having DRAM and more about having a decent controller and firmware.

The NM790 is a known good drive that is priced well and has a decent track record. I'm using DRAMless WD SN7100s for most builds at the moment and they are indistinguishable in real-world use from systems that have SN850X and Samung 980 Pros in them with DRAM.

TPU's SSD reviews prove that the difference between a good SSD and a great SSD are so small that you'll struggle to even measure them most of the time, so a 'good' SSD like the NM790 is plenty for just about anyone, and simply having DRAM does not guarantee that the drive is suitable as a prosumer/workstation-grade drive.
 
I've just abandoned four really cheap DRAM-less Lexar NQ100/NS100 SATA SSDs as Windows boot drives, because they were slowing down after only 4 days use and grinding to a halt after 9 days. One 120GB, two 240GB and one 480GB. All 2.5" laptop size.

I know I should be using more Samsung 870 EVOs with DRAM, but these are old systems I use for test purposes and I'm not prepared to buy a dozen EVOs for obsolete systems. Only the most important machines get EVOs.

When I ran Hard Disk Sentinel's Surface read test on the 9-day old drive, it ploughed through the bad areas on the SDD at 17MB/s. After running HD Sentinel's destructive Write test, the bad areas (marked with dark green squares on the disk map) all disappeared and the subsequent read test ran at 460MB/s.

I have never experienced such rapid slow down with similarly cheap SATA SSDs from the likes of Patriot (Burst Elite, P210 and P220), Pny and Integral. I've had 120GB DRAM-less boot drives running for years and they're still going strong, apart from the Lexars. My main PCs contain an eclectic mix of Samsung, Kingston, Crucial and WD M.2 NVMe drives, but I won't be buying any more Lexar drives.

The SATA Lexars will go in my junk box until Lexar release new firmware.

As it happens, I also had problems in 2024 with a brand new 128GB Lexar Compact Flash card. After a few tests, it stopped working so I sent it back and paid double the price for two Sandisk 64GB CF cards. Serves me right for trying to save money.
 
I've just abandoned four really cheap DRAM-less Lexar NQ100/NS100 SATA SSDs as Windows boot drives, because they were slowing down after only 4 days use and grinding to a halt after 9 days. One 120GB, two 240GB and one 480GB. All 2.5" laptop size.

I know I should be using more Samsung 870 EVOs with DRAM, but these are old systems I use for test purposes and I'm not prepared to buy a dozen EVOs for obsolete systems. Only the most important machines get EVOs.

When I ran Hard Disk Sentinel's Surface read test on the 9-day old drive, it ploughed through the bad areas on the SDD at 17MB/s. After running HD Sentinel's destructive Write test, the bad areas (marked with dark green squares on the disk map) all disappeared and the subsequent read test ran at 460MB/s.

I have never experienced such rapid slow down with similarly cheap SATA SSDs from the likes of Patriot (Burst Elite, P210 and P220), Pny and Integral. I've had 120GB DRAM-less boot drives running for years and they're still going strong, apart from the Lexars. My main PCs contain an eclectic mix of Samsung, Kingston, Crucial and WD M.2 NVMe drives, but I won't be buying any more Lexar drives.

The SATA Lexars will go in my junk box until Lexar release new firmware.

As it happens, I also had problems in 2024 with a brand new 128GB Lexar Compact Flash card. After a few tests, it stopped working so I sent it back and paid double the price for two Sandisk 64GB CF cards. Serves me right for trying to save money.
DRAMless SATA is a different beast, they are dreadful, since SATA doesn't support HMB so you're left with a shit drive that performs like garbage with no DRAM to use at all.

DRAMless isn't a problem for NVMe drives because HMB works just as well as on-drive DRAM, and the on-drive DRAM is usually super-cheap, super-slow stuff like DDR4-2133. It's not like there's a huge advantage to having DRAM on the SSD when your system ram is likely 2-3x faster. Yes, the latency is marginally higher than on-drive DRAM, but NAND latency is many orders of magnitude slower, so it really doesn't matter where the DRAM is, and it's why manufacturers use the cheapest, slowest DRAM they can find still in production for their drives.
 
Using same drive and had no issues like this. It may be bad sample.
 
Using same drive and had no issues like this. It may be bad sample.
Pretty sure OP just broke their PCIe slot and the problem was 100% unrelated to the Lexar NM790.
This thread could have been closed by post #12.
 
You live and you learn. We've all been there at some point. Glad you got things sorted & working!

I've just abandoned four really cheap DRAM-less Lexar NQ100/NS100 SATA SSDs as Windows boot drives, because they were slowing down after only 4 days use and grinding to a halt after 9 days. One 120GB, two 240GB and one 480GB. All 2.5" laptop size.

I know I should be using more Samsung 870 EVOs with DRAM, but these are old systems I use for test purposes and I'm not prepared to buy a dozen EVOs for obsolete systems. Only the most important machines get EVOs.

When I ran Hard Disk Sentinel's Surface read test on the 9-day old drive, it ploughed through the bad areas on the SDD at 17MB/s. After running HD Sentinel's destructive Write test, the bad areas (marked with dark green squares on the disk map) all disappeared and the subsequent read test ran at 460MB/s.

I have never experienced such rapid slow down with similarly cheap SATA SSDs from the likes of Patriot (Burst Elite, P210 and P220), Pny and Integral. I've had 120GB DRAM-less boot drives running for years and they're still going strong, apart from the Lexars. My main PCs contain an eclectic mix of Samsung, Kingston, Crucial and WD M.2 NVMe drives, but I won't be buying any more Lexar drives.

The SATA Lexars will go in my junk box until Lexar release new firmware.

As it happens, I also had problems in 2024 with a brand new 128GB Lexar Compact Flash card. After a few tests, it stopped working so I sent it back and paid double the price for two Sandisk 64GB CF cards. Serves me right for trying to save money.

Using same drive and had no issues like this. It may be bad sample.

Pretty sure OP just broke their PCIe slot and the problem was 100% unrelated to the Lexar NM790.
This thread could have been closed by post #12.
Well, I spoke too soon. I had to wait few days until I got a new thermal paste. I plugged everything into the right places, but nah, it is still being slow and laggy!!!!! What on earth is going on?!?!

Here is what my fancontrol says as well as taskmanager:
327412638126.jpg.47284adf7f80c31760bdf259a90c44b6.jpg

And this is the cable for the AIO, the thin black one:
IMG_20250724_205717.jpg

Also, the AIO (Liquid Freezer II) has started to make "liquid noise" now! Listen to this video (crank up the volume): https://streamable.com/e1pw09

So. I am stuck now. I do not know what else to do... Is the AIO not working? The cpu that has gone bonkers?

ANY help would be kindly appreciated!
 
AIO maybe borked heating up the CPU and slowing everything down….you do monitor temps when using it?
 
Why isn't your motherboard screwed in? four of the four holes you've taken a photo of don't have screws in them.

Alignment of the GPU and the motherboard is really really important to avoid weird PCIe lane issues, which you've obviously already had so far in this thread along with the broken PCIe latch not locking the GPU into its correct alignment in the slot and now I've spotted that your board isn't even anchored securely in place. Is your GPU the only thing holding your entire motherboard upright?! Also, while it's not necessary in a healthy system, the motherboard has grounding points that connect to the motherboard screws and they're not just for decoration. Get some screws in them so that your board's ground rails are properly grounded.

Start by fixing the really really simple stuff before moving onto more complex and harder-to-diagnose troubleshooting! :)
 
Back
Top