Intel is keeping Optane/3D XPoint, selling only the NAND part of their memory business to SK Hynix. The sale is structured so that Intel still has use of the NAND factory until 2025, so it seems while Intel might not see want to be in the NAND business long term, in the short to mid term they still have plans.
They're still reliable. I rarely see Intel SSD's fail - super common in OEM applications for this reason. But they gave up the performance crown years ago. A decade ago.
Anand wrote a great piece on Intel toward the end of his tenure here at AT (I think it was around 2013) describing Intel's intention with their SSD business. They are, after all, a company that must please investors and turn a profit. And Intel has always been profit first. That said, the consumer SSD business was never a priority for them. The only reason they were even in it is because the storage mediums of yesteryear (spinning rust) were impeding their ability to sell CPU's; the user experience was more affected by hard disk performance than CPU performance.
So Intel jumpstarted the consumer SSD business with the X25 series drives. They were solid, reliable, fast, and relatively cheap for the time. After that, they pretty much gave up, because their job was done once competitors came in with mainstream drives.
I believe that message carries on to this day. Some of their decisions are highly questionable (I still don't understand the purpose of Optane, but seeing how Intel is balls deep into QLC, perhaps Intel believe Optane will be necessary to hide the appalling QLC performance delta.
Optane has little significance for consumers with being other medium cache is the top use as of right now. Optane for data centers has more uses for its specific qualities.
"They're still reliable. I rarely see Intel SSD's fail"
Depends on the definition of failure. The latency of the 660p is pretty fail in some books. The other drawbacks are also failures in that, no matter how much they're papered-over by "SLC" cache and the like — they still exist.
The only thing QLC has going for it is 30% density. How far will we chase diminishing density returns? PLC? HLC?
We can't build decent NAND on leading-edge processes, even if the capacity was available. There's only so many things they can do to drive up bits per dollar. There's a growing demand for more capacity for the money. I don't know how far they'll go beyond QLC, but QLC NAND isn't going away any time soon.
With that being said, I still use MLC for my main drive, and QLC for secondary storage only.
fairly sure first-gen Optane did not perform as well as intel hoped for. The biggest problem for them was that they could not max pcie(3) 4x speeds like Nand manufacturers could while still costing more. Add there a fact that it seems that there is a bottleneck somewhere else. Since game load times do not significantly go down even when using ramdisk I am fairly sure the issue is somewhere between how windows reads files and how games or game-engine uses that interface. If CPU was the issue it would be utilized 100% when reading files ( which is not the case at least for me) if it was because of disks using ramdisk should show greater improvement. Sadly for intel Optane for storage falls into this pit where other people's screwups make it kinda useless unless you are in server environment where request queue always has plenty of things to do. Hopefully this bottleneck gets fixed at some point and we can enjoy even lower load times.
Is it just me, or are the QLC haters nuts? Look, I get that the price needs to come down compared to TLC. But there's a simple fact that is going to lead to QLC taking over the consumer end at some point: the more cells there are, the less frequently those cells get rewritten. A 16TB future QLC drive has a much higher TBW than any current TLC drive.
Why are you commenting on a topic you know nothing about? QLC has just 33% more space than TLC, but several times worse write endurance. Doesn't take a hater to see it's just a bad deal.
I would argue that most people will benefit more from the extra storage of QLC instead of the higher endurance of TLC/MLC (Let's be real here - how likely is it that you will blow through the >100TBW rating of modern SSDs?).
QLC makes more sense for things like media file storage, where you write once and read many times. For heavy regular use TLC makes better sense, but MLC makes more sense then that and so on.
I totally agree, I I were and enthusiast user I'd go with MLC with SLC for win, TLC for games and QLC for data instead of HDD. And this is what I did 8 years ago with the 840 pro. I use my pc for work, media tasks, and now I want something more, that's why I like optane
This is incorrect. TLC = 3 bits per cell, or 8 states. QLC = 4 bits per cell, or 16 states. QLC = 2X the data density of TLC. I.E. 200% more data per cell, not 33%. It's not 3 states to 4 states, it's 3 bits to 4 bits per cell, which is huge.
@ceiver, no, it's 33% more bits equal 33% more storage. You don't measure capacity by the number of 'states' it can hold. Otherwise a 2TB drive would in actuality be hugely bigger than 1TB having twice the number of bits for <insert ginormous number> more states.
Reflex brings up a good point. QLC will be perfected much like TLC was once it went 3D. I don't know what technological innovation it will take for QLC to become fast and long-term reliable (like more than 70 drive writes before imminent failure) but there will inevitably be something. As of now, they hide the performance issues with QLC using crazy SLC caches or Optane cache drives, much like the storage cartels hide SMR performance issues with a NAND cache. Neither of those solutions is very good (especially with SMR drives I think the NAND should be 8-16GB not 1-2GB) but it's what they have right now and it will improve.
If we're talking strictly "retail" availability then it's been ~2.5 years for QLC, and V-NAND just refers to vertically stacked designs. It took Samsung 3 years to move between mass-manufacture of initial TLC designs and the first 3D NAND TLC with reasonable performance; handling 16 voltage levels per cell is significantly more difficult than 8, so it stands to reason that any potential improvement in design will take longer to shake out of the tree.
That said, I do doubt it's going to suddenly improve in any drastic way. I strongly suspect that improvements will have to be more indirect than physical transformations of the NAND itself.
I think it's people's natural nervousness. They are always worried about having enough of a limited resource. Without a clear knowledge or understanding of what they need and what is available this nervousness really kicks in. Some people have this nervousness more than others.
Nobody cares when somebody first announced a cell design that was techically capable of reading out 16 voltage levels. The fact that it took 5-6 years longer than TLC to hit the market indicates how much more difficult it was to turn that into a usable product - you're comparing apples to oranges. All you're really proving with those dates is that if QLC is going to improve, it will not be a fast process.
QLC does not have more cells. QLC uses 4 bits per cell, which means each flash memory transistor supports writing/reading 16 distinct voltage levels (charge amounts stored in the floating gate of the transistor). Therefore in QLC flash each cell gets written more often than each cell in TLC memory of the same capacity.
You're missing the point entirely. By your logic, we should hate TLC because it's less endurant than MLC at the same capacity. But we don't. Why? There is a threshold at which endurance starts to matter, and that threshold is constantly being pushed back by capacity gains in all drive types. So I'll say it again: none of you QLC haters are worried about your TLC drive, but a future QLC drive will have a HIGHER TBW rating than your current TLC drive, because the endurance loss per cell is totally negated by having that many more cells to spread writes across.
But you don't get more cells with QLC! Why is this so hard to grasp? The number of cells stays the same, and the capacity increase is due to each cell storing more data. Going from 3 bit per cell to 4 bit per cell you get 33% capacity increase, so each cell is now 33% less likely to be used, but at the same time it's 100% more likely to be used due to it's storing 1 more bit. As a result, each cell is now more likely to be used, and it will get even worse when we go from 4 bit to 5 bit.
On the one hand, I understand people's irritation that QLC still hasn't provided a significant cost saving over older TLC drives despite being an inferior option in every regard except capacity.
On the other hand, yes, some people's insistence on astroturfing every comment section on articles related to NAND with the same incessant bitching and trash-talk is getting very, very old.
Those h20 drives... just give me BOTH capacities as 2 drives in single slot. This could be so useful as swap space for + normal ssd for laptops that mostly ride a single slot....
did I miss something or is this the first time that Intel admits to Ice Lake being a 2021 product? I think they called Q4 even in the Q2 investors call
I think that during the October 22nd Q3 earnings call Intel said that Ice Lake servers would begin shipping in early 2021. I think that was the time they announced the delay.
it seems to moi that QLC in the Enterprise Space actually makes more sense than retail. here's why: Enterprise don't run a drive until it dies, they swap out bad ones as they happen or end of warrant period whichever comes first, and if too many die before warranty, they switch suppliers. the added capacity of QLC is a win, modulo bad supply. how many Enterprise really, really need the fastest random I/O on the planet? precious few, I'd wager. what the want is cheap capacity, in minimal footprint, minimal power draw, and better than some version of spinning rust (I know, it's not iron-oxide any more). retail, OTOH, understands (incorrectly) that solid state devices, should they survive infant death, last for-freakin-ever. QLC, without planned and executed replacement, isn't such a good deal.
But a QLC SSD would use less power than an SSHD would. It's not even a little bit, it's going to be several watts per drive, which across an entire datacenter is an enormous power increase/savings (depending on which way you're going)...
"retail, OTOH, understands (incorrectly) that solid state devices, should they survive infant death, last for-freakin-ever."
SSDs have warranties on them. I had a Kingston SSD (don't blame me, I didn't buy it) die on me after a couple of years of light use. Sent it in under warranty, got sent a new one no problem.
Very nice to see Alder Stream (2nd gen Optane) finally being released! I can't wait it to use it in some virtual hosts. This should allow for even higher performance and/or higher virtual guest counts.
At least on the 900p Intel shipped a M.2 to U.2 adapter with the drive at retail. AFAIK it's totally passive, just wires connecting a card to the U.2 connector. It works fine, though routing sucks a bit because of the power insertion in the middle. There are a couple of no-name companies making similar adapters as well.
Alder Stream SSD specs are phenomenal with 100 DWPD and insane speed. I hope we see client version released once Rocket lake cpu's are released. There is no reason even AMD desktops cannot use optane ssd. I hope ultimately we see optane ssd for high end laptops as well(not optane H20 crap).
It is a shame there is no enthusiast/workstation M2 version of the second gen 3d xpoint. I hope they will release something after 590 chipset, otherwise I would buy the datacenter version with an adapter. 3D xpoint is the perfect solution I think for workstation main NVME, incredible endurance and perfomances. At the time I bought the Samsung 840 pro and It was a perfect experience for years, consistent performance well over its endurance time. I also bought 850, 860, 870 evo as secondary drives and the experience is nowhere near the mighty pro. Considering the new Top of the line nvme now have TLC memory I will gladly spend 10 times more to have a new Optane
Optane failed as anything with intel lately. Where are the 128-512GB optane m.2s for the user? Would be a nice OS/page file never worrying about writes
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
69 Comments
Back to Article
Chaitanya - Wednesday, December 16, 2020 - link
Surprising announcement given Intel recently announced sale of this division.ilt24 - Wednesday, December 16, 2020 - link
@ChaitanyaIntel is keeping Optane/3D XPoint, selling only the NAND part of their memory business to SK Hynix. The sale is structured so that Intel still has use of the NAND factory until 2025, so it seems while Intel might not see want to be in the NAND business long term, in the short to mid term they still have plans.
FreckledTrout - Thursday, December 17, 2020 - link
Also the deal isn't done. Its not expected to close until end of 2021.powerarmour - Wednesday, December 16, 2020 - link
Intel SSD's at least used to be good. Too much QLC now.MetaCube - Wednesday, December 16, 2020 - link
QLC bad, please upvote now69369369 - Wednesday, December 16, 2020 - link
qLcBADDDDD haha im so original haha le reddit moment +3 gold +40 silver +8 wholesome +39 keanuchungus69369369 - Wednesday, December 16, 2020 - link
I'm sorry.TheinsanegamerN - Wednesday, December 16, 2020 - link
You speak the truthBeaver M. - Thursday, December 17, 2020 - link
When you think youre cool hating on haters because you think you deserve to be cooler than them because you hate them.Spunjji - Friday, December 18, 2020 - link
When you think you're cool hating on hater-haters because you think you deserve to be cooler than them because you hate them.Beaver M. - Friday, December 18, 2020 - link
When you think you're cool hating on hater-haters-haters because you think you deserve to be cooler than them because you hate them.Beaver M. - Friday, December 18, 2020 - link
Jokes aside, even if you dont want to admit it, I think you got the point I was trying to make. Comments like his are stupid, to put it generously.69369369 - Friday, December 18, 2020 - link
That's the idea. (Tommy Wiseau laugh) :-)azfacea - Thursday, December 17, 2020 - link
we all hate qlc as much we hated IE back when ms was forcing everyone 2 buy it as part of windows. ms paid for that, and so will these evil doersSpunjji - Friday, December 18, 2020 - link
"evil doers"Okay, but tell us what you really feel about NAND.
Samus - Wednesday, December 16, 2020 - link
They're still reliable. I rarely see Intel SSD's fail - super common in OEM applications for this reason. But they gave up the performance crown years ago. A decade ago.Anand wrote a great piece on Intel toward the end of his tenure here at AT (I think it was around 2013) describing Intel's intention with their SSD business. They are, after all, a company that must please investors and turn a profit. And Intel has always been profit first. That said, the consumer SSD business was never a priority for them. The only reason they were even in it is because the storage mediums of yesteryear (spinning rust) were impeding their ability to sell CPU's; the user experience was more affected by hard disk performance than CPU performance.
So Intel jumpstarted the consumer SSD business with the X25 series drives. They were solid, reliable, fast, and relatively cheap for the time. After that, they pretty much gave up, because their job was done once competitors came in with mainstream drives.
I believe that message carries on to this day. Some of their decisions are highly questionable (I still don't understand the purpose of Optane, but seeing how Intel is balls deep into QLC, perhaps Intel believe Optane will be necessary to hide the appalling QLC performance delta.
Eliadbu - Thursday, December 17, 2020 - link
Optane has little significance for consumers with being other medium cache is the top use as of right now. Optane for data centers has more uses for its specific qualities.Oxford Guy - Monday, December 21, 2020 - link
"They're still reliable. I rarely see Intel SSD's fail"Depends on the definition of failure. The latency of the 660p is pretty fail in some books. The other drawbacks are also failures in that, no matter how much they're papered-over by "SLC" cache and the like — they still exist.
The only thing QLC has going for it is 30% density. How far will we chase diminishing density returns? PLC? HLC?
Alexvrb - Friday, January 1, 2021 - link
We can't build decent NAND on leading-edge processes, even if the capacity was available. There's only so many things they can do to drive up bits per dollar. There's a growing demand for more capacity for the money. I don't know how far they'll go beyond QLC, but QLC NAND isn't going away any time soon.With that being said, I still use MLC for my main drive, and QLC for secondary storage only.
compvter - Monday, January 4, 2021 - link
fairly sure first-gen Optane did not perform as well as intel hoped for. The biggest problem for them was that they could not max pcie(3) 4x speeds like Nand manufacturers could while still costing more. Add there a fact that it seems that there is a bottleneck somewhere else. Since game load times do not significantly go down even when using ramdisk I am fairly sure the issue is somewhere between how windows reads files and how games or game-engine uses that interface. If CPU was the issue it would be utilized 100% when reading files ( which is not the case at least for me) if it was because of disks using ramdisk should show greater improvement. Sadly for intel Optane for storage falls into this pit where other people's screwups make it kinda useless unless you are in server environment where request queue always has plenty of things to do. Hopefully this bottleneck gets fixed at some point and we can enjoy even lower load times.euler007 - Tuesday, January 5, 2021 - link
Holy crap, it's already been 6+ years since he left, felt like yesterday.arslan0123 - Tuesday, December 22, 2020 - link
QLC should cost 50% less but this black friday, samsung Evo and QLC same price in Canada, many people bought without knowing about itbansheexyz - Wednesday, December 16, 2020 - link
Is it just me, or are the QLC haters nuts? Look, I get that the price needs to come down compared to TLC. But there's a simple fact that is going to lead to QLC taking over the consumer end at some point: the more cells there are, the less frequently those cells get rewritten. A 16TB future QLC drive has a much higher TBW than any current TLC drive.benedict - Wednesday, December 16, 2020 - link
Why are you commenting on a topic you know nothing about? QLC has just 33% more space than TLC, but several times worse write endurance. Doesn't take a hater to see it's just a bad deal.69369369 - Wednesday, December 16, 2020 - link
I would argue that most people will benefit more from the extra storage of QLC instead of the higher endurance of TLC/MLC (Let's be real here - how likely is it that you will blow through the >100TBW rating of modern SSDs?).Oxford Guy - Monday, December 21, 2020 - link
Endurance is not the only problem with QLC.TheinsanegamerN - Wednesday, December 16, 2020 - link
QLC makes more sense for things like media file storage, where you write once and read many times. For heavy regular use TLC makes better sense, but MLC makes more sense then that and so on.umano - Friday, December 18, 2020 - link
I totally agree, I I were and enthusiast user I'd go with MLC with SLC for win, TLC for games and QLC for data instead of HDD. And this is what I did 8 years ago with the 840 pro. I use my pc for work, media tasks, and now I want something more, that's why I like optaneceiver - Wednesday, December 16, 2020 - link
This is incorrect. TLC = 3 bits per cell, or 8 states. QLC = 4 bits per cell, or 16 states. QLC = 2X the data density of TLC. I.E. 200% more data per cell, not 33%. It's not 3 states to 4 states, it's 3 bits to 4 bits per cell, which is huge.FwFred - Wednesday, December 16, 2020 - link
@ceiver, no, it's 33% more bits equal 33% more storage. You don't measure capacity by the number of 'states' it can hold. Otherwise a 2TB drive would in actuality be hugely bigger than 1TB having twice the number of bits for <insert ginormous number> more states.Spunjji - Friday, December 18, 2020 - link
That's some bad mathsReflex - Wednesday, December 16, 2020 - link
You are correct. We've heard this paranoia on every change from SLC to MLC to TLC and now QLC. It's never been a problem.Samus - Wednesday, December 16, 2020 - link
Reflex brings up a good point. QLC will be perfected much like TLC was once it went 3D. I don't know what technological innovation it will take for QLC to become fast and long-term reliable (like more than 70 drive writes before imminent failure) but there will inevitably be something. As of now, they hide the performance issues with QLC using crazy SLC caches or Optane cache drives, much like the storage cartels hide SMR performance issues with a NAND cache. Neither of those solutions is very good (especially with SMR drives I think the NAND should be 8-16GB not 1-2GB) but it's what they have right now and it will improve.Beaver M. - Thursday, December 17, 2020 - link
QLC was first introduced in 2009 and is already 3 years in retail as V-NAND!Theres not going to be a "super improved" QLC.
Spunjji - Friday, December 18, 2020 - link
If we're talking strictly "retail" availability then it's been ~2.5 years for QLC, and V-NAND just refers to vertically stacked designs. It took Samsung 3 years to move between mass-manufacture of initial TLC designs and the first 3D NAND TLC with reasonable performance; handling 16 voltage levels per cell is significantly more difficult than 8, so it stands to reason that any potential improvement in design will take longer to shake out of the tree.That said, I do doubt it's going to suddenly improve in any drastic way. I strongly suspect that improvements will have to be more indirect than physical transformations of the NAND itself.
Oxford Guy - Monday, December 21, 2020 - link
If anything, it's only going to get worse as nodes shrink.Yojimbo - Wednesday, December 16, 2020 - link
I think it's people's natural nervousness. They are always worried about having enough of a limited resource. Without a clear knowledge or understanding of what they need and what is available this nervousness really kicks in. Some people have this nervousness more than others.drexnx - Wednesday, December 16, 2020 - link
the performance lost for the miniscule (consumer) cost savings achieved makes QLC an absolute joke.TheinsanegamerN - Wednesday, December 16, 2020 - link
Show me all those 8TB TLC drives then. Fas as I know only samsung ever made one, and it was stupid expensive.TLC has been around for some time, QLC is brand new. TLC had no price advantage when it first came out either.
Beaver M. - Thursday, December 17, 2020 - link
QLC brand new? QLC is 3 years old and was first developed in 2009!Spunjji - Friday, December 18, 2020 - link
Nobody cares when somebody first announced a cell design that was techically capable of reading out 16 voltage levels. The fact that it took 5-6 years longer than TLC to hit the market indicates how much more difficult it was to turn that into a usable product - you're comparing apples to oranges. All you're really proving with those dates is that if QLC is going to improve, it will not be a fast process.Oxford Guy - Monday, December 21, 2020 - link
"Show me all those 8TB TLC drives then"Why should companies sell us those when they can make more profit selling QLC instead?
shelbystripes - Wednesday, December 16, 2020 - link
What do you mean by "miniscule"? I'm waiting to see these cheap-as-QLC 4TB and 8TB TLC SSDs... where are they?p1esk - Wednesday, December 16, 2020 - link
QLC does not have more cells. QLC uses 4 bits per cell, which means each flash memory transistor supports writing/reading 16 distinct voltage levels (charge amounts stored in the floating gate of the transistor). Therefore in QLC flash each cell gets written more often than each cell in TLC memory of the same capacity.bansheexyz - Thursday, December 17, 2020 - link
You're missing the point entirely. By your logic, we should hate TLC because it's less endurant than MLC at the same capacity. But we don't. Why? There is a threshold at which endurance starts to matter, and that threshold is constantly being pushed back by capacity gains in all drive types. So I'll say it again: none of you QLC haters are worried about your TLC drive, but a future QLC drive will have a HIGHER TBW rating than your current TLC drive, because the endurance loss per cell is totally negated by having that many more cells to spread writes across.p1esk - Friday, December 18, 2020 - link
But you don't get more cells with QLC! Why is this so hard to grasp? The number of cells stays the same, and the capacity increase is due to each cell storing more data. Going from 3 bit per cell to 4 bit per cell you get 33% capacity increase, so each cell is now 33% less likely to be used, but at the same time it's 100% more likely to be used due to it's storing 1 more bit. As a result, each cell is now more likely to be used, and it will get even worse when we go from 4 bit to 5 bit.shelbystripes - Wednesday, December 16, 2020 - link
The QLC haters are definitely nuts. And misinformed. And doomed to a lifetime of disappointment.Spunjji - Friday, December 18, 2020 - link
On the one hand, I understand people's irritation that QLC still hasn't provided a significant cost saving over older TLC drives despite being an inferior option in every regard except capacity.On the other hand, yes, some people's insistence on astroturfing every comment section on articles related to NAND with the same incessant bitching and trash-talk is getting very, very old.
deil - Wednesday, December 16, 2020 - link
Those h20 drives... just give me BOTH capacities as 2 drives in single slot.This could be so useful as swap space for + normal ssd for laptops that mostly ride a single slot....
dersteffeneilers - Wednesday, December 16, 2020 - link
did I miss something or is this the first time that Intel admits to Ice Lake being a 2021 product? I think they called Q4 even in the Q2 investors callYojimbo - Wednesday, December 16, 2020 - link
I think that during the October 22nd Q3 earnings call Intel said that Ice Lake servers would begin shipping in early 2021. I think that was the time they announced the delay.
Spunjji - Friday, December 18, 2020 - link
Stretching the definition of "announced" a little there :DFunBunny2 - Wednesday, December 16, 2020 - link
it seems to moi that QLC in the Enterprise Space actually makes more sense than retail. here's why: Enterprise don't run a drive until it dies, they swap out bad ones as they happen or end of warrant period whichever comes first, and if too many die before warranty, they switch suppliers. the added capacity of QLC is a win, modulo bad supply. how many Enterprise really, really need the fastest random I/O on the planet? precious few, I'd wager. what the want is cheap capacity, in minimal footprint, minimal power draw, and better than some version of spinning rust (I know, it's not iron-oxide any more). retail, OTOH, understands (incorrectly) that solid state devices, should they survive infant death, last for-freakin-ever. QLC, without planned and executed replacement, isn't such a good deal.dersteffeneilers - Wednesday, December 16, 2020 - link
I mean hard drives + RAM caching are right thereTheinsanegamerN - Wednesday, December 16, 2020 - link
Which onyl makes sense for certian use cases. SSDs fill roles where that wouldnt help at all.shelbystripes - Wednesday, December 16, 2020 - link
But a QLC SSD would use less power than an SSHD would. It's not even a little bit, it's going to be several watts per drive, which across an entire datacenter is an enormous power increase/savings (depending on which way you're going)...Tomatotech - Thursday, December 17, 2020 - link
"retail, OTOH, understands (incorrectly) that solid state devices, should they survive infant death, last for-freakin-ever."SSDs have warranties on them. I had a Kingston SSD (don't blame me, I didn't buy it) die on me after a couple of years of light use. Sent it in under warranty, got sent a new one no problem.
romrunning - Wednesday, December 16, 2020 - link
Very nice to see Alder Stream (2nd gen Optane) finally being released! I can't wait it to use it in some virtual hosts. This should allow for even higher performance and/or higher virtual guest counts.RealBeast - Thursday, December 17, 2020 - link
Yeah, that was my real takeaway here. 2nd Gen Optane is getting a lot more interesting, now just to get the prices down to Earthly levels.imaskar - Wednesday, December 16, 2020 - link
No successor to Optane M10? That's a bummer.M15 was announced, but never released...
Pinn - Wednesday, December 16, 2020 - link
Any consumer mobos have a U.2 port?NonSequitor - Wednesday, December 16, 2020 - link
At least on the 900p Intel shipped a M.2 to U.2 adapter with the drive at retail. AFAIK it's totally passive, just wires connecting a card to the U.2 connector. It works fine, though routing sucks a bit because of the power insertion in the middle. There are a couple of no-name companies making similar adapters as well.scrizz - Monday, December 21, 2020 - link
yes, this is one example. there are moreEVGA Z390/Z490 DARK
trivik12 - Wednesday, December 16, 2020 - link
Alder Stream SSD specs are phenomenal with 100 DWPD and insane speed. I hope we see client version released once Rocket lake cpu's are released. There is no reason even AMD desktops cannot use optane ssd. I hope ultimately we see optane ssd for high end laptops as well(not optane H20 crap).umano - Friday, December 18, 2020 - link
It is a shame there is no enthusiast/workstation M2 version of the second gen 3d xpoint. I hope they will release something after 590 chipset, otherwise I would buy the datacenter version with an adapter. 3D xpoint is the perfect solution I think for workstation main NVME, incredible endurance and perfomances. At the time I bought the Samsung 840 pro and It was a perfect experience for years, consistent performance well over its endurance time. I also bought 850, 860, 870 evo as secondary drives and the experience is nowhere near the mighty pro. Considering the new Top of the line nvme now have TLC memory I will gladly spend 10 times more to have a new Optaneedzieba - Monday, December 21, 2020 - link
Ugh, I hadn't noticed how awful that logo was before (or is it new branding?)! What happened to OPTAN 'a' through 'd'?!catavalon21 - Sunday, January 17, 2021 - link
If you're referring to the lower case "e" following all-caps, apparently it's now a thing.https://www.intel.com/content/www/us/en/homepage.h...
Lolimaster - Sunday, December 27, 2020 - link
Optane failed as anything with intel lately. Where are the 128-512GB optane m.2s for the user? Would be a nice OS/page file never worrying about writesDeicidium369 - Monday, December 28, 2020 - link
Optane is for the data center - and has hardly failed.