• Guest, The rules for the P & N subforum have been updated to prohibit "ad hominem" or personal attacks against other posters. See the full details in the post "Politics and News Rules & Guidelines."

Do you think we will see U.2 hard drives?

cbn

Lifer
Mar 27, 2009
12,968
221
106
Perhaps ones with the following techs or mix of the following techs:

1. SSHD, either with NAND or 3D Xpoint
2. Dual drive (SSD and Hard drive combined into the same form factor)
3. Internal RAID, apparently called Redundant array independent platters--> https://community.spiceworks.com/topic/254904-hard-drives-with-internal-raid?page=1 (However, instead of the use mentioned in the link a 4TB hard drive with four 1TB platters (as an example) could be set-up in a fashion similar to having four 1TB hard drives in RAID 0.....which would boost sequential read and write)
 

FFFF

Member
Dec 20, 2015
199
18
36
The only use case I can think of are Optane-HDD hybrids like those announced in the new Thinkpad line, but I doubt it.
 

cbn

Lifer
Mar 27, 2009
12,968
221
106
Some more info on that here.

So Lenovo will have 16GB version of Optane (in the form of M.2 2242 SSD) along with a HDD as an alternative to a single NVMe SSD in the soon to be released Thinkpad models.

Lenovo's announcement today of a new generation of ThinkPads based on Intel's Kaby Lake platform includes brief but tantalizing mention of Optane, Intel's brand for devices using the 3D XPoint non-volatile memory technology they co-developed with Micron. Lenovo's new ThinkPads and competing high-end Kaby Lake systems will likely be the first appearance of 3D XPoint memory in the consumer PC market.
Several of Lenovo's newly announced ThinkPads will offer 16GB Optane SSDs in M.2 2242 form factor paired with hard drives as an alternative to a using a single NVMe SSD with NAND flash memory (usually TLC NAND, with a portion used as SLC cache). The new Intel Optane devices mentioned by Lenovo are most likely the codenamed Stony Beach NVMe PCIe 3 x2 drives that were featured in roadmap leaked back in July. More recent leaks have indicated that these will be branded as the Intel Optane Memory 8000p series, with a 32GB capacity in addition to the 16GB Lenovo will be using. Since Intel's 3D XPoint memory is being manufactured as a two-layer 128Gb (16GB) die, these Optane products will require just one or two dies and will have no trouble fitting on to a short M.2 2242 card alongside a controller chip.
 
Last edited:
  • Like
Reactions: stewart497

stewart497

Junior Member
Jan 15, 2017
4
2
1
Some more info on that here.

So Lenovo will have 16GB version of Optane (in the form of M.2 2242 SSD) along with a HDD as an alternative to a single NVMe SSD in the soon to be released Thinkpad models.
I wonder how much it will cost?)
 

Insert_Nickname

Diamond Member
May 6, 2012
4,392
923
136
Perhaps ones with the following techs or mix of the following techs:

1. SSHD, either with NAND or 3D Xpoint
2. Dual drive (SSD and Hard drive combined into the same form factor)
I don't suppose its impossible. There are plenty of SAS drives already using the connector, it is just a matter of using PCIe instead.

3. Internal RAID, apparently called Redundant array independent platters--> https://community.spiceworks.com/topic/254904-hard-drives-with-internal-raid?page=1 (However, instead of the use mentioned in the link a 4TB hard drive with four 1TB platters (as an example) could be set-up in a fashion similar to having four 1TB hard drives in RAID 0.....which would boost sequential read and write)
...? o_O

That is how HDDs work already. You can already achieve "RAID-on-disk" by keeping your most frequently accessed data (think swap file) on the outermost cylinders. An old school performance trick is making a small partion at the start of the disk (outer cylinders, hence fastest), and assigning your swap file to it. Did this way back in the nineties... :D
 

cbn

Lifer
Mar 27, 2009
12,968
221
106
Some more info on that here.

So Lenovo will have 16GB version of Optane (in the form of M.2 2242 SSD) along with a HDD as an alternative to a single NVMe SSD in the soon to be released Thinkpad models.
I wonder how much it will cost?)
Optane is supposed to cost about 5x more than than NAND on a GB per GB per basis (ie, so 16GB Optane is around the cost of 80GB NAND).

With that mentioned, I wonder what the hard drive capacity will be? The current Thinkpad models (with the exception of the T460s) have a 500GB 2.5" 7mm:

http://shop.lenovo.com/us/en/laptops/thinkpad/t-series/

..............but I am going to guess these Thinkpad models with the 16GB optane cache would have at least a 1TB 2.5" 7mm (probably 2TB 2.5" 7mm) since the soon to be released Thinkpads are also offering up to a 512GB NVMe SSD.
 

Insert_Nickname

Diamond Member
May 6, 2012
4,392
923
136
Not so sure. HDD head-assemblies move together, with the individual head active and used to fine-tune the seek.

They do not have multiple head assemblies that move independently.
True, the R/W heads move at the same time, but data is striped across the platters. Sectors and tracks line up on each platter, and each set of R/W heads work independently. Assuming 4 platters you get the first sector written on the first platter, second on the second, third on third, forth on forth. Then repeat until the first track is full. Only then does the headassembly move inwards a notch to the next track. It is not like each individual platter is filled first, and then moving on to the next one.

At least that is how I remember it. Could be wrong.
 

cbn

Lifer
Mar 27, 2009
12,968
221
106
I found in the following Article Seagate Sandforce having SATA 6Gbps and NVMe combined in one controller (for both Client and Enterprise):

http://www.anandtech.com/show/9362/seagate-announces-sandforce-sf3500-ssd-controller-series-mass-production-expected-in-q415

I am wondering if they thought about also using this for SATA 6Gbps SSHD in the beginning.....and then eventually transition the SSHD over to PCIe for both the solid state drive part of the drive and the platters? It does have four channels though (and DRAM buffer)....which does seem a bit much for a controller typically used with SSHD. However, if they were (or are) planning to have SSHDs with greater amounts of NAND than we typically see today then I could see such a controller making sense.
 

Valantar

Golden Member
Aug 26, 2014
1,792
508
136
I seriously doubt this will ever happen. HDDs don't even come close to maxing out SATA 2, let alone 3, so adding bandwidth is a pure waste. Also, PCIe lanes are at a premium in pretty much every system out there, while SATA lanes are not. Not to mention that a modular approach would work better, having an Optane (or similar) chache off-drive allows for it to be modular and replaceable, allowing for upgrades or larger capacities serving multiple HDDs.
 

cbn

Lifer
Mar 27, 2009
12,968
221
106
Doing some more research I did find out that SATA Express does support PCIe 3.0 x 2, not just PCIe 2.0 x 2.

So that is pretty decent and could make for some good SSHDs.

However, just to be clear SATA Express and U.2 are quite different from each other.

P.S. Back in 2014, Western Digital did release a PCIe Hard drive. It was a dual drive though, not a SSHD:

http://www.storagereview.com/wd_demonstrates_first_pcie_hard_drives

 
Last edited:

Valantar

Golden Member
Aug 26, 2014
1,792
508
136
Doing some more research I did find out that SATA Express does support PCIe 3.0 x 2, not just PCIe 2.0 x 2.

So that is pretty decent and could make for some good SSHDs.

However, just to be clear SATA Express and U.2 are quite different from each other.

P.S. Back in 2014, Western Digital did release a PCIe Hard drive. It was a dual drive though, not a SSHD:

http://www.storagereview.com/wd_demonstrates_first_pcie_hard_drives

That picture is a nice demonstration of why SATA Express died a silent death.
 

thecoolnessrune

Diamond Member
Jun 8, 2005
9,603
510
126
That picture is a nice demonstration of why SATA Express died a silent death.
They are still putting it on Motherboards, even the affordable B250 chipset board below:

https://www.newegg.com/Product/Product.aspx?Item=N82E16813128984&cm_re=GA-B250M-D3H-_-13-128-984-_-Product
.
SATA Express isn't going anywhere right now, as it's one of the best options in the Consumer market for general storage. That's why AMD is equipping it on their new motherboards. It is extremely flexible. It is backwards compatible with the current SATA connector. It can be one Sata Express port, or two SATA Ports. It can also simply be two PCI-e 3.0 lanes (which is more than enough for a mainstream PCI-e SSD). Unlike M.2, you can easily take a drive off-board, which is great considering how restricted the M.2 form factor is. Heat and NAND space are constant constraints.

U.2 is great (I'm a huge fan), but it will face the same issue as SAS. U.2 is designed to accommodate dual controllers, which no general consumer really needs / is willing to pay the cost for. SATA Express on the other hand will be great in the future I feel for allowing decent performance on large, slow TLC 2.5 or even 3.5" SSDs. I can imagine people using their "high speed" M.2 SSDs for OS and critical apps, while bulk storage will slowly be moving towards 2-4TB bulk TLC drives. They'll be comparatively slow compared to an M.2 drive, but they'll still take decent advantage of a PCI-e 2x interface. In the mean time, people don't have to decide to lose PCI-e lanes on M.2 slots or lose SATA ports.

And even again, if the motherboard manufacturer doesn't want a board that supports any of that stuff, they can remove the option entirely and get 4 general purpose PCI-e lanes.

I think AMD including SATA Express was a fantastic move for flexibility.
 

Ichinisan

Lifer
Oct 9, 2002
28,298
1,232
136
3. Internal RAID, apparently called Redundant array independent platters--> https://community.spiceworks.com/topic/254904-hard-drives-with-internal-raid?page=1 (However, instead of the use mentioned in the link a 4TB hard drive with four 1TB platters (as an example) could be set-up in a fashion similar to having four 1TB hard drives in RAID 0.....which would boost sequential read and write)
Pretty sure they already work kinda like that. Hard drive engineers the last 20 years aren't dumb.
 

Valantar

Golden Member
Aug 26, 2014
1,792
508
136
SATA Express isn't going anywhere right now, as it's one of the best options in the Consumer market for general storage. That's why AMD is equipping it on their new motherboards. It is extremely flexible. It is backwards compatible with the current SATA connector. It can be one Sata Express port, or two SATA Ports. It can also simply be two PCI-e 3.0 lanes (which is more than enough for a mainstream PCI-e SSD). Unlike M.2, you can easily take a drive off-board, which is great considering how restricted the M.2 form factor is. Heat and NAND space are constant constraints.

U.2 is great (I'm a huge fan), but it will face the same issue as SAS. U.2 is designed to accommodate dual controllers, which no general consumer really needs / is willing to pay the cost for. SATA Express on the other hand will be great in the future I feel for allowing decent performance on large, slow TLC 2.5 or even 3.5" SSDs. I can imagine people using their "high speed" M.2 SSDs for OS and critical apps, while bulk storage will slowly be moving towards 2-4TB bulk TLC drives. They'll be comparatively slow compared to an M.2 drive, but they'll still take decent advantage of a PCI-e 2x interface. In the mean time, people don't have to decide to lose PCI-e lanes on M.2 slots or lose SATA ports.

And even again, if the motherboard manufacturer doesn't want a board that supports any of that stuff, they can remove the option entirely and get 4 general purpose PCI-e lanes.

I think AMD including SATA Express was a fantastic move for flexibility.
I haven't seen a single AM4 board with SATA Express (I admit I haven't been looking hard, but I've seem many, many boards lacking it). Also, there doesn't exist a single SATAe storage device in the market, nor does any storage device maker have any of these on their currently publicized roadmaps - some did a few years ago, and all have been cancelled. While the connector might be flexible, it's also stillborn. It will never see adoption. Its PCIe x2 connectivity is seen as too much of a limitation for performance when competing with m.2's x4, while the horribe tri-ribbon cable is both wildly impractical to route, almost impossible to buy, and requires drives to have power connectors on a weird appendage cable thingy. No thank you.

Edit: I looked over Asus and ASrock's selections of AM4 motherboards. Not a single one has a SATA Express connector.


U.2 is rather impractical for consumer use, as it makes little sense over m.2. Still, for consumer usage, it serves purely as a "m.2 over a cable" interface. It might see some limited use, but that's unlikely too. The only real reason for it to see adoption is lack of board space for more m.2 slots. Also, talking about "large, slow TLC 2.5 or 3.5" SSDs" makes no sense. If they're large due to a large number of flash chips, this makes them both expensive (flash is expensive) and fast (parallelism). If they're large due to high-density chips, that makes them slow (lack of parallelism), but still not cheap. Adding the cost of a PCIe (even x2) controller won't help, nor will limiting the drive's audience to those with SATAe connectors. Drives like the former will get PCIe x4 connections, as they'll easily saturate an x2 connection on sequential reads (parallelism, again), and there's nearly zero added cost for this. Drives like the latter work perfectly fine on SATA.

Oh, and an SSD with enough PCB space and flash dies to require a 3.5" housing will inevitably be in the very very very high price range.


I'm sorry, but your entire post is completely out of touch with reality.
 
  • Like
Reactions: VirtualLarry

corkyg

Elite Member | Peripherals
Super Moderator
Mar 4, 2000
27,370
237
106
eSATA may be dead, but the one in my Thinkpad is terrific - USB3 speeds with better reliability. :)
 
  • Like
Reactions: Hi-Fi Man

thecoolnessrune

Diamond Member
Jun 8, 2005
9,603
510
126
I haven't seen a single AM4 board with SATA Express (I admit I haven't been looking hard, but I've seem many, many boards lacking it). Also, there doesn't exist a single SATAe storage device in the market, nor does any storage device maker have any of these on their currently publicized roadmaps - some did a few years ago, and all have been cancelled. While the connector might be flexible, it's also stillborn. It will never see adoption. Its PCIe x2 connectivity is seen as too much of a limitation for performance when competing with m.2's x4, while the horribe tri-ribbon cable is both wildly impractical to route, almost impossible to buy, and requires drives to have power connectors on a weird appendage cable thingy. No thank you.

Edit: I looked over Asus and ASrock's selections of AM4 motherboards. Not a single one has a SATA Express connector.


U.2 is rather impractical for consumer use, as it makes little sense over m.2. Still, for consumer usage, it serves purely as a "m.2 over a cable" interface. It might see some limited use, but that's unlikely too. The only real reason for it to see adoption is lack of board space for more m.2 slots. Also, talking about "large, slow TLC 2.5 or 3.5" SSDs" makes no sense. If they're large due to a large number of flash chips, this makes them both expensive (flash is expensive) and fast (parallelism). If they're large due to high-density chips, that makes them slow (lack of parallelism), but still not cheap. Adding the cost of a PCIe (even x2) controller won't help, nor will limiting the drive's audience to those with SATAe connectors. Drives like the former will get PCIe x4 connections, as they'll easily saturate an x2 connection on sequential reads (parallelism, again), and there's nearly zero added cost for this. Drives like the latter work perfectly fine on SATA.

Oh, and an SSD with enough PCB space and flash dies to require a 3.5" housing will inevitably be in the very very very high price range.


I'm sorry, but your entire post is completely out of touch with reality.
It's really hard to take you seriously (especially the "out of touch with reality" part), when the #2 selling motherboard manufacturer (Gigabyte) has no less than 3 motherboards at launch with SATA Express. Maybe you'd be a bit more in touch with reality if you looked a little harder, because I would say if you couldn't find those boards, you're not looking hard at all.

The great thing about SATA Express is that it does not require SATA Express to be successful in Drive factor, or even exist. Having the option gives the OEM total control over the end product. If they want, the end user can deploy 2 SATA ports. If they want, the end user can deploy 2 PCI-e NVM drives. If they want, the OEM can take the 4 PCI-e lanes and redirect that for any general purpose they want, or just make a slot. Tell me how M.2 would fill that role? U.2? Anything? Remember most boards with SATA Express also have M.2 Drive slots. This isn't a replacement for the space constricted M.2, it's a replacement for the protocol restricting SATA port. Again, any rational argument for why the port shouldn't be there would be fine.

The rest of your point regarding flash drives is complete conjecture no different from my own. You make giant logical leaps that somehow, a drive with NAND that isn't necessarily the fastest, will reap no benefit from lower latency from NVMe, as well as increased parallelism. Enterprise ultra-high capacity drives will continue to slowly lower the price of storage in brackets underneath them.

The SSD market at this point grows roughly 9% per year, and while not linear, SSD drives have decreased in price roughly 85% compared to the start of the decade.

If you have any studied data that indicates the expected prices of Solid State Drives over the next 3 years, as well as their expected performance profiles, which would add value to your point by indicating that drive density / performance / and prices would never leave any exposure for a pure SATA Express solution, I'd be interested in that. Otherwise, acknowledging that it's your assumption, just as mine was, is alright too. It's better than the "out of touch with reality" crap you're using now.
 

cbn

Lifer
Mar 27, 2009
12,968
221
106
Looking into this a bit further, I notice Puget Systems describes B250 as being able to have either one M.2 (PCIe 3.0 x 4) SSD or two SATA Express (PCIe 3.0 x 2) drives:

https://www.pugetsystems.com/labs/articles/Z270-H270-Q270-Q250-B250---What-is-the-Difference-876/

So one large NVMe M.2 SSD (using either 3D TLC or 3D QLC) vs. one small or medium size NVMe M.2 SSD + SATA 6 Gbps Hard drives (or SSDs) vs two SATA Express SSHD in RAID-0?

(Or basically, looking at the extreme end of things, one large M.2 (or U.2 via adapter) NVMe with 3D QLC vs. two SATA Express SSHD in RAID-0?)
 
Last edited:

Valantar

Golden Member
Aug 26, 2014
1,792
508
136
It's really hard to take you seriously (especially the "out of touch with reality" part), when the #2 selling motherboard manufacturer (Gigabyte) has no less than 3 motherboards at launch with SATA Express. Maybe you'd be a bit more in touch with reality if you looked a little harder, because I would say if you couldn't find those boards, you're not looking hard at all.
Last generation (Z170 in particular) had SATAe on essentially every board from every manufacturer. That Gigabyte only has it on three is a sign of diminishing implementations.

The great thing about SATA Express is that it does not require SATA Express to be successful in Drive factor, or even exist. Having the option gives the OEM total control over the end product. If they want, the end user can deploy 2 SATA ports. If they want, the end user can deploy 2 PCI-e NVM drives. If they want, the OEM can take the 4 PCI-e lanes and redirect that for any general purpose they want, or just make a slot. Tell me how M.2 would fill that role? U.2? Anything? Remember most boards with SATA Express also have M.2 Drive slots. This isn't a replacement for the space constricted M.2, it's a replacement for the protocol restricting SATA port. Again, any rational argument for why the port shouldn't be there would be fine.
While I agree with you in theory, the practical end result is that it's nearly impossible to make a NVMe drive for PCIe x2 that couldn't possibly benefit from x4, and the controller and development cost is pretty much identical. As such, all drives going forward are made with an x4 interface, and thus will never get SATAe connectors as those would place an artificial limitation on drive performance.

Also, you've got your numbers wrong. SATAe is two SATA ports or two lanes of PCIe 2.0/3.0, allowing for a single drive to use these lanes. If you're talking about the common dual SATAe connector blocks, those have four SATA ports, two independent SATAe ports, and the use of one SATAe doesn't affect the availability of the other SATA ports.

The rest of your point regarding flash drives is complete conjecture no different from my own. You make giant logical leaps that somehow, a drive with NAND that isn't necessarily the fastest, will reap no benefit from lower latency from NVMe, as well as increased parallelism. Enterprise ultra-high capacity drives will continue to slowly lower the price of storage in brackets underneath them.
It might be conjecture, but it's supported by pretty much every tech writer in the world. The added cost of a PCIe controller pretty much necessitates an x4 interface to eke out as much performance as possible to recoup as much of the development cost as possible. This applies whether the drive is meant for low-end or high-end applications. Besides, Intel is already making the 600p, which is barely more expensive than SATA SDDs. As such, there is no logical reason to go for the in-between solution of an x2 interface - either go whole hog x4, or SATA with the significant cost savings this brings with it. There is definitely a market for low-end NVMe drives, just like there is a market for high-end SATA drives. I just don't see any reason to gimp these drives with an x2 interface that will further limit their performance.

The SSD market at this point grows roughly 9% per year, and while not linear, SSD drives have decreased in price roughly 85% compared to the start of the decade.

If you have any studied data that indicates the expected prices of Solid State Drives over the next 3 years, as well as their expected performance profiles, which would add value to your point by indicating that drive density / performance / and prices would never leave any exposure for a pure SATA Express solution, I'd be interested in that. Otherwise, acknowledging that it's your assumption, just as mine was, is alright too. It's better than the "out of touch with reality" crap you're using now.
I'm not saying it's impossible that a pure SATAe drive will appear at some point. I'm just saying that based on three simple facts - that manufacturers have abandoned/canceled all known SATAe device development, motherboard manufacturers are moving away from the standard, and there's no extra cost involved in using m.2/u.2 instead - there's no reason to suspect this will ever happen.

As for rational arguments against SATAe: it adds board complexity, it's a horribly designed connector (seriously, that thing is awful), and device manufacturers have shown no desire to utilize it.

I would love a way to easily hook up x2 PCIe over a cable - that has a million possible use cases. The problem is that SATAe is a bad solution to this problem. While its innate flexibility is nice in principle, in practice it makes for a huge, ungainly and impractical connector with a nightmarish cable. U.2 takes up less board space, does more, and its size makes whether or not users use them just as moot a point as with the SATAe connector. The ideal way to solve this would be u.2 ports with support for lane splitting. Much, much more convenient than SATAe.
 

cbn

Lifer
Mar 27, 2009
12,968
221
106
(snip) the practical end result is that it's nearly impossible to make a NVMe drive for PCIe x2 that couldn't possibly benefit from x4, and the controller and development cost is pretty much identical. As such, all drives going forward are made with an x4 interface.
The PCIe 3.0 x 2 controllers are definitely cheaper.

The following is an excerpt from a past Anandtech article explaining this:

On the PCIe side JMicron has canceled the JMF810 and JMF811 controllers, and will now be focusing solely on the JMF815. JMicron made the decision to concentrate on the value segment and thus the JMF815 is a PCIe 3.0 x2 design with four NAND channels (no NVMe, unfortunately). A four-lane design would have required moving to 28nm process node, which would have increased the cost substantially and the packaging would have to move away from BGA to FCBGA (used by e.g. Phison and SandForce in their upcoming PCIe controllers) that would further increase the cost. I think it's a good play from JMicron to focus on a segment that isn't as populated because right now everyone is focusing solely on performance with PCIe, but ultimately cost and power consumption will be a major factors in widespread adoption and JMicron should have an advantage there if the JMF815 is executed well.
P.S. The PCIe 3.0 x 2 controller mentioned in the link I provided is not NVMe, but Marvell 88NV1160 and Phison E8/E8T are NVMe PCIe 3.0 x 2 controllers. (Samsung also makes a NVMe PCIe 3.0 x 2 controller called "Photon")
 

thecoolnessrune

Diamond Member
Jun 8, 2005
9,603
510
126
Last generation (Z170 in particular) had SATAe on essentially every board from every manufacturer. That Gigabyte only has it on three is a sign of diminishing implementations.
That seriously sounds like a complete guess. How many motherboards had SATA Express at launch? How many AM4 motherboards are out there now for a lunch platform? Do you believe every motherboard following this AM4 launch will be feature parity with existing boards throughout the service life? You're taking a bunch of observations (new platform, new CPUs, new chipset, new designs) and saying that it's diminishing, which it is, but you provide complete conjecture as to *why* its diminishing. That also includes the fact that you seem to think it has some sort of "permanence" about it. Samsung SSD shipments were down Q1 2016, but no one considered it permanent. And it wasn't.


While I agree with you in theory, the practical end result is that it's nearly impossible to make a NVMe drive for PCIe x2 that couldn't possibly benefit from x4, and the controller and development cost is pretty much identical. As such, all drives going forward are made with an x4 interface, and thus will never get SATAe connectors as those would place an artificial limitation on drive performance.
Again, complete conjecture, and absolutely wrong. Companies like JMICRON have already noted scrapping x4 PCIe controllers in exchange for x2 controllers to hit price points.

Also, you've got your numbers wrong. SATAe is two SATA ports or two lanes of PCIe 2.0/3.0, allowing for a single drive to use these lanes. If you're talking about the common dual SATAe connector blocks, those have four SATA ports, two independent SATAe ports, and the use of one SATAe doesn't affect the availability of the other SATA ports.
My numbers aren't wrong, but I could see the mis-interpretations. In my original post, I was discussing AM4, because AMD is implementing SATA Express on Chipset. In that area, reviewers have noted that the 2 SATA Express Ports can be divided into SATA, used as 2 x2 PCIe NVMe ports, or coalesced into an x4 PCIe port, hence why it can also be another M.2 slot.


It might be conjecture, but it's supported by pretty much every tech writer in the world. The added cost of a PCIe controller pretty much necessitates an x4 interface to eke out as much performance as possible to recoup as much of the development cost as possible. This applies whether the drive is meant for low-end or high-end applications. Besides, Intel is already making the 600p, which is barely more expensive than SATA SDDs. As such, there is no logical reason to go for the in-between solution of an x2 interface - either go whole hog x4, or SATA with the significant cost savings this brings with it. There is definitely a market for low-end NVMe drives, just like there is a market for high-end SATA drives. I just don't see any reason to gimp these drives with an x2 interface that will further limit their performance.
Again, as you noted, it's conjecture. You have provided no evidence, while I have provided evidence indicating the opposite: http://www.anandtech.com/show/9369/jmicron-ssd-controller-roadmap-jmf680-sata-6gbps-jmf815-pcie-controllers-next-year

Controllers stay on process nodes a very long time due to cost reasons. Even after the jump to 28nm, it still doesn't necessarily mean there aren't room for x2 controllers. You're welcome to bring conjecture if you want, but stating it as some sort of fact is silly.

EDIT: Noting that @cbn posted this information ahead of me. I made this post in my spare time over 3 hours, so I missed it. :)


I'm not saying it's impossible that a pure SATAe drive will appear at some point. I'm just saying that based on three simple facts - that manufacturers have abandoned/canceled all known SATAe device development, motherboard manufacturers are moving away from the standard, and there's no extra cost involved in using m.2/u.2 instead - there's no reason to suspect this will ever happen.
To the first point, that's fine. Again, SATA Express does not require devices be made for the interface to live.
To the second, a single brand new platform launch does not a trend make. Since you said "facts", can you bring any evidence forward with any numbers that shows SATA Express is releasing on fewer motherboards within stable platforms? Or is this more of an observation?

As for rational arguments against SATAe: it adds board complexity, it's a horribly designed connector (seriously, that thing is awful), and device manufacturers have shown no desire to utilize it.
Again, compared to what. You have yet to provide a valid alternative. It adds board complexity? How? AM4 leaves it entirely up to the OEM. They could not implement it at all, or they could make additional SATA ports. Or they could even make the block a x4 PCIe slot, or hell, make it an M.2 slot. The idea behind AM4's implementation is that it gives the OEM total freedom on the final implementation without the complexity of having to add an additional controller to the motherboard. How can M.2 help with that? How can U.2 help with that? Why do you think that space constricted M.2 slots do not add to board complexity when you have to route a slot, and make sure no components interfere with the M.2 card that ends up mounted on he motherboard. Seriously, that doesn't make any sense. As for the connector, compared to what? M.2 only has a kludge of a connector. U.2 carries 2 channels as well but is a much more robust cable, SFF-8639. The cables are individually shielded! Have you ever bought Mini HD SAS cables? Are you aware of how much those things cost right now?

I would love a way to easily hook up x2 PCIe over a cable - that has a million possible use cases. The problem is that SATAe is a bad solution to this problem. While its innate flexibility is nice in principle, in practice it makes for a huge, ungainly and impractical connector with a nightmarish cable. U.2 takes up less board space, does more, and its size makes whether or not users use them just as moot a point as with the SATAe connector. The ideal way to solve this would be u.2 ports with support for lane splitting. Much, much more convenient than SATAe.
I would love that too. Lots of uses. But we don't have it. The rest is simply an opinion which I can't agree with. U.2 takes less board space as a connector, but gives up flexibility. It's entirely personal opinion on which you would rather have. I agree the SATA Express cable is therefore much larger, but that's a trade-off. Do you think people will want to spend $40 on what's essentially a SAS3 cable? We can't even get people to spend more than $70 on their power supplies a great deal of the time.

U.2 lane splitting already exists, because it's primarily a commercial product designed for Enterpise use. Dual Port U.2 drives like the Intel DC3600 have been on the market for a year now. But they go many to one, rather than the many to many you're wanting. That's because U.2 either sends 4x PCIe for 1 controller, or 2, 2x PCIe for dual controller mode. Unlike SATA Express, all of the lanes are sent down each U.2 connector to each device. How would you do what you wanted? Would you go back to the days of master / slave drives where you manually set each drive, or use a separate cable that only had half the lanes in each U.2 connector? That standard doesn't exist by the way. You'd have to create a whole new SFF cable standard. And even when you did achieve that, you'd end up with 2, 2x PCI-e connectors. That seems like a lot of work for something we already have though. It's called SATA Express ;)
 

VirtualLarry

No Lifer
Aug 25, 2001
52,589
7,261
126
SATA-E is stillborn. Let it go, man. :p

I would love for it to evolve into a nice simple connector, like SATA, except it would have x4 PCI-E lanes, to run the same hardware controllers that are currently used on M.2 drives, except in a small external form-factor with a cable. Why? Because current M.2 standards aren't efficient with board space. Then again, there's the U.2 connector, so maybe that will take off instead.

But SATA-E, as it is NOW, is effectively dead. No mfgs have even announced any shipping products for it. And when Z170 was introduced with SATA-E, it was on basically ALL motherboards that supported it in the chipset. No mfg held back. Now they are holding back.
 
  • Like
Reactions: Valantar

thecoolnessrune

Diamond Member
Jun 8, 2005
9,603
510
126
SATA-E is stillborn. Let it go, man. :p

I would love for it to evolve into a nice simple connector, like SATA, except it would have x4 PCI-E lanes, to run the same hardware controllers that are currently used on M.2 drives, except in a small external form-factor with a cable. Why? Because current M.2 standards aren't efficient with board space. Then again, there's the U.2 connector, so maybe that will take off instead.

But SATA-E, as it is NOW, is effectively dead. No mfgs have even announced any shipping products for it. And when Z170 was introduced with SATA-E, it was on basically ALL motherboards that supported it in the chipset. No mfg held back. Now they are holding back.
VL, SATA Express stillborn as a device endpoint does not necessarily mean that the interface is going away. It's low on deployment for the AM4 platform, but I'm fine with the option being available. If it dies, it dies. It makes no real difference for me. But what I dislike is this fanciful interpretation of history and computing. You're doing the same things.

You mentioned U.2. Again. Did you read my previous post fully? Do you think end users will be fine with purchasing $40-$70 cables to connect to their SSDs? Valantar wanted the option to multi-lane as well, which already exists, as the end result is SATA Express.

Seriously, I was looking for rational arguments for, in the current world, with what we have, why the interface is bad. Thus far, the only thing I've heard with real merit is that the "connector is large".

It seems a lot of you guys could get what you want by having 2 U.2 connectors + a PCI-e Mux. You'd still be bottle-necked by x4 upstream, but you'd have x4 bandwidth to each port, which would work well if you didn't heavily load each end device. So we're talking 2 $40 cables, and the add-on cost of the Mux switch (plus the value-add cost to you as a feature).

How much extra on top of current board cost would you be willing to pay for your desires to be real?
 

ASK THE COMMUNITY