As I hinted during last week's podcast, I am in Seoul, Korea to cover an upcoming press conference.
To those keen readers who have followed my previous trips here, it can only mean one thing –
..and with a Samsung SSD Global Summit comes product announcements. Those don't happen until tomorrow (late tonight for you folks back in the states), but I did notice a clue on the cover of our itinerary folder:
See it? Here, let me help:
A VNAND powered M.2 (presumably NVMe) SSD is *exactly* the thing I have been waiting for Samsung to unleash into the wild ever since we reviewed their NVMe SM951. Given that Samsung's prior M.2 offerings gave the Intel SSD 750 a run for its money all while consuming half the power, and did so with Samsung's older 2D Planar NAND, you can bet a VNAND version will be something to behold. Let's hope this new model is released as a consumer product and doesn't end up as OEM-channel unobtanium like the NVMe SM951 was!
Keep an eye out for additional posts from our coverage of the 2015 Samsung SSD Global Summit!
CRAP! I just bought an
CRAP! I just bought an SM951!!! Oh well, I have budget to spare for a new one…
WHOOO BLACK PCB!!!! Yeah,
WHOOO BLACK PCB!!!! Yeah, lets hope it’s not made of OEM Super-Unobtainium :/
That chip could fit nicely in
That chip could fit nicely in your jeans pocket Allyn. Then maybe a give-away? >.>;
Dude, get us some photos of
Dude, get us some photos of the city too please. At night and day time. The residential areas and the main city areas too! 🙂
Specs ASAP please. And don’t
Specs ASAP please. And don’t forget to smuggle one back for testing lol.
The bare M.2 without a
The bare M.2 without a heatsink is a little worrying. The SM951 had a lot of heat/throttling issues.
They were claiming 30% power
They were claiming 30% power savings for their new 48-layer devices, so this may be better in that respect. Anyway, to achieve maximum performance in a desktop machine, it would be best to go with an actual full PCI-e card. There isn’t that much of a market for such high speed in the consumer space though. There are some enthusiast who would buy it, but it is probably massive overkill for consumer applications.
We didn’t see any throttling
We didn't see any throttling in our testing of the SM951's, and that was with *very little* air flow across the PCB. It really only heated up when hitting it with sustained writes (for several minutes straight), which is a very rare case for a consumer SSD.
Did people put it in laptops
Did people put it in laptops with zero airflow perhaps? In just about any kind of desktop, I would think it would have sufficient airflow.
Did people put it in laptops
Did people put it in laptops with zero airflow perhaps? In just about any kind of desktop, I would think it would have sufficient airflow.
Another must-have piece of
Another must-have piece of equipment is a cell phone to
enable easy response to your customers. Will you
need the cleaning service to do a thorough cleaning for you.
Equipment should also be serviced on a weekly basis to make them more efficient.
my weblog :: شركة نقل عفش بالدمام, queens-movers.net,
Your not afraid of flying.
Your not afraid of flying. The only way I’d ever get to South Korea is by boat.
Query: If you’re given a
Query: If you’re given a single M.2 PCIe 2.0 x2 slot (say, on an Asus Z97-A motherboard), and you’re given an M.2 drive built for M.2 PCIe 2.0 x2 and one built for PCIe 3.0 x4, would they perform identically/similarly? (As in, the 3.0 x4 drive, limited to 2.0 x2, would only be strangled to performance similar to native 2.0 x2 drives?)
The only real limit change
The only real limit change would be in sequential throughput. Latency would still be very low and random performance would be similar (up until you hit the lower bandwidth cap). You do give up some straight line speed but it's mostly the same experience.
I wonder if anyone will make
I wonder if anyone will make a PCI-e raid card that just has m.2 slots on it. It doesn’t seem like it would be that complicated since it is mostly just a PCI-e switch coupled to a raid controller. I don’t know how expensive it would be to have something like 4 x4 downstream and one x8 or x16 upstream.
It’s trickier than you think,
It's trickier than you think, especially with NVMe devices. You really need driver level support (and also need BIOS support if you plan to boot from it). More information on that in our Z170 Skylake PCIe storage performance review.
I doubt such a set-up (PCI-e
I doubt such a set-up (PCI-e switch) would be supported by Intel RST RAID; that isn’t really what I was thinking of. Intel RST seems quite a bit more flexible than what I expected, but I didn’t expect to be able to RAID PCI-e drives at all, at least not this soon. The consumer space doesn’t really have any need for such RAID support anyway. Even a single PCI-e ssd is probably overkill for most consumers right now. It seems like a hardware (not “firmware” or software) RAID card with m.2 slots would be an interesting product for some enterprise use though. It would pack a lot of really high-speed storage into a very small form factor using mostly commodity parts which would also be individually replaceable.
This has been a long time
This has been a long time coming, hopefully I can find one and without it breaking the bank.
They of course will initially
They of course will initially break all but the most robust banks. HDD speed is still the bottleneck for most work, so these huge leaps in speed (or are they just evolutions?) will be sucked into the mainstream fairly quickly. Any system will benefit from a hard disk speed increase, not just bleeding edge systems, so expect to see these move around quickly.
To be honest more interested
To be honest more interested in 2,4,6+ TB AHCI SSDs than NVMe from Samsung, but will take whatever comes our way.
I hope they can keep on
I hope they can keep on scaling up the number of layers. They may have a chance of competing with hard drive cost per GB if they can. Although, that may need to be combined with 4-bit cells.