SK Hynix Sampling New PCIe 4.0 96L SSDs, 128L 4D NAND Enterprise SSDs
by Andrei Frumusanu on April 8, 2020 11:00 AM ESTSK Hynix today announced that they’ve begun sampling of its first ever PCIe 4.0 enterprise SSDs in the form of the new 96-layer 3D-NAND U.2/U.3 form-factor PE8010 and PE8030 eSSDs, as well as announcing plans to sample the new PE81111 EDSFF E1.L SSDs based on their 128-layer “4D NAND” flash modules later in the year.
We had expected the new PE8111 eSSD for some time know as we reported about SK Hynix’s plans to introduce such a product last November. The biggest change here is the company’s use of new 128-layer 3D NAND modules that the company dubs as “4D-NAND” because of a new denser cell structure design and higher per-die I/O speeds.
16TB Enterprise EDSFF E1.L SSD
The PE8111 still retains as PCIe 3.0 interface and its corresponding performance characteristics plateau at 3400MB/s sequential reads and 3000MB/s sequential writes – whilst supporting random reads and writes up to respectively 700K and 100K IOPs. Because it’s a long-factor EDSFF E1.L form-factor, storage capacity for the unit falls in at 16TB, and SK Hynix is reporting that they’re working on a 32TB solution in the future.
The new PE8010 and PE8030 come in an U.2/U.3 form-factor and are the company’s first SSDs support PCIe 4.0. The SSDs here still rely on 96-layer NAND modules from the company – but are using an in-house controller chip. Bandwidth here is naturally higher, reaching up to 6500MB/s reads and 3700MB/s write sequentially, with random IOPs falling in at respectively 1100K for reads and 320K for writes.
Power consumption for the new U.2/U.3 drives is actually extremely competitive given their jump to PCIe 4.0 – rising only up to 17W as opposed to their previous generation PCIe 3.0 products which fell in at 14W. This is likely to be attributed to the new generation custom controller, which might be more optimised for low-power compared some or the early third-party 4.0 controllers out there.
The PE8010 and PE8030 are sampling right now with customers – with the PE8111 planned to be sampled in the second half of the year.
Related Reading:
- Enterprise NVMe Round-Up 2: SK Hynix, Samsung, DapuStor and DERA
- SK Hynix to Cut CapEx, Accelerate Transitions, 1z nm DRAM & 128L 4D NAND in 2020
- SK Hynix Sampling 128-layer 3D NAND SSDs
- SK Hynix NAND Update: 3D NAND Output Cut, Slowdown Capacity Expansions
- SK Hynix Starts Production of 128-Layer 4D NAND, 176-Layer Being Developed
Source: SK Hynix
40 Comments
View All Comments
azfacea - Thursday, April 9, 2020 - link
and whats the math pcie 4 and 5 SSD VS ddr4 and ddr5could u also do the math on GB per dollar, max GB, TB, ... in a 1u, 2u, ....
include some raid setups in your TCO.
Kristian Vättö - Friday, April 10, 2020 - link
Bandwidth is only one part of the equation. In majority of workloads latency is far more critical because that is what the end-user actually experiences. There is still an order of magnitude difference between DRAM (10's of nanoseconds) and SSD/NAND (10's of microseconds). No interface is going to solve the latency issue because it boils down to the fundamental physics of DRAM and NAND architectures.rhysiam - Saturday, April 11, 2020 - link
Thank you @Kristian. This. Just this!All the hype around 3D Xpoint was primarily because it had a latency that was - from the marketing marketing hype - approaching that of DRAM. But even XPoint latency, which is far superior to any NAND on the market, is still way too high to replace DRAM for general purpose computing. The idea that NAND could replace DRAM based on iterations of current technology is is nonsense.
kn00tcn - Wednesday, April 8, 2020 - link
you're so focused on your obnoxious replies that you can't even see the fact that ssds are mostly the exact same dram companies with the exact same price fixing wavesazfacea - Wednesday, April 8, 2020 - link
what u say is patently false. 6.5 + China insurgency is not at all the same as 3 entrenched. we've seen how diff is x86 with just one insurgent.obnoxious indeed. cant see facts indeed.
TheinsanegamerN - Friday, April 10, 2020 - link
Meanwhile here comes DDR5 with 2-3X the bandwidth of DDR4.This is like saying "if mobile GPUs keep getting better there will be no reason to have desktop GPUs". Its factually incorrect. DRAM will always be faster then NAND, and you will always have a need for both. They are different types of memory for different applications.
azfacea - Thursday, April 9, 2020 - link
if anand tech don't like the continuity between story and comments they can always make comments a toggle able section or use discuss or something. I am using comments exactly as the mainstream does, and not abusing them in any special manner.I am very skeptic of "group think". and have strong opinions which are mostly right but also sometimes wrong. that does lead to confrontatious threads sometimes. if u can't bear that get out of the kitchen. go watch sesame street instead.
when ppl attack with condescending dismissals implying my ignorance, and w/- facts BTW, I am going get back at them
schujj07 - Friday, April 10, 2020 - link
Just because you think you are right doesn't mean you are. The people who have responded to you aren't using "group think" to dismiss your point. The fact that you are asking that we provide hard facts to counter your "opinion" but don't provide any hard evidence yourself shows ignorance on your part. Multiple people have no provided you with evidence to counter your point but you don't listen. That puts us in an impossible position where no matter what we say, how much evidence we provide, etc... in your mind we will always be wrong and you will always be right. Don't forget that the people who are countering your argument and opinions, like @Kristian Vättö, & myself work in the IT field. For myself I deal with these things on a daily basis and have studied it for years.schujj07 - Friday, April 10, 2020 - link
Edit: Multiple people have NOW provided you with evidence to counter your point but you don't listen.azfacea - Thursday, April 9, 2020 - link
here is where the negativity in this thread started and its not me:i said "competitive pressure" and not replace DRAM with SSD, he replied:
" Even the fastest SSD are several orders of magnitude slower than DRAM. "
in other words i am that ignorant. he didnt even bother to read or think about my comment. just an impulsive pushback that i am wrong and ignorant. dont have have a clue, and if only I wasnt ignorant of DRAM performance numbers I'd know better than to say that.
In other words not a counter argument, just an education reply is what he gave. and w/- facts BTW. he is dead wrong, I am right, with this kinda of performance growth for SSDs it truly means they are begging to compete with DRAM in at least some applications. and Xbox/PS is perfect example.
he started with a condescending dismissal, and i gave him a taste of his own medicine with the network comment, i could have phrased it differently but why should I? when he is being like that.
now what does he do, turn it into an argument about the messenger. and so does the rest of "group think" echo chamber