News ASRock Unveils the X299E-ITX/ac: Mini ITX + X299 + Quad-channel Memory

Mod edit:



Detailed overview of what we know about the X299E-ITX/ac thus far here: https://smallformfactor.net/news/asrock-x299e-itxac-little-monster-detailed

Original:

ASRock did it! Finally, there's an Intel HEDT platform motherboard with full quad-channel DDR4 memory. The new X299E-ITX/ac is for those who need up to 18 CPU cores and up to 64 GB of quad-channel DDR4 memory in their SFF machines for reasons. The board manages its limited PCB real-estate by going vertical. It features two riser cards, one with a few onboard controllers, and a pair of 32 Gb/s M.2 slots), and the other riser with SATA 6 Gb/s ports, a third M.2 slot, and the headers such as USB 3.1. The board draws power from 24-pin ATX and 8-pin EPS connectors, conditioning it for the LGA2066 CPU using a 7-phase VRM. The lone expansion slot is a PCI-Express 3.0 x16, memory is handled by four DDR4 SO-DIMM slots. Connectivity includes two Intel I219-V driven gigabit Ethernet interfaces, 802.11ac WLAN, and Bluetooth 4.1.



Source
 
Last edited by a moderator:

BirdofPrey

Standards Guru
Sep 3, 2015
797
493
I've seen that sort of dongle business before with storage as well.
Some SAS RAID adapters require them to unlock certain premium features )the main one that I've seen being SSD caching)

It's a shame Intel seems to be going even more all in with their market segmentation. With Intel's reaction to Ryzen and Threadripper being to slash prices on their own HEDT line maybe in a year or two AMD offering SMT on their entire product line not cutting PCIe lanes out of their low cost parts, and letting Threadripper and Epyc share the same platform (at least as far as we know), maybe we can hope that Intel might follow suit and relax their artificial restraints a bit.

Hmm, maybe ASRock can work their BIOS magic and get around that.
You mean like how various mainboard manufacturers have unlocked non-K processors through their BIOS only to have Intel update their CPU firmware to again disable it?
 

jeshikat

Jessica. Wayward SFF.n Founder
Silver Supporter
Feb 22, 2015
4,969
4,781
You mean like how various mainboard manufacturers have unlocked non-K processors through their BIOS only to have Intel update their CPU firmware to again disable it?

Right, ASRock was real big on that. But hopefully Intel wouldn't patch it out because limiting it to Intel SSDs is just totally stupid.
 

BirdofPrey

Standards Guru
Sep 3, 2015
797
493
Yeah, but they do that sort of stuff a LOT.
Part of their artificial market segmentation. I mentioned it elsewhere, but I am hoping increased competition from AMD makes that a less viable strategy.

Beyond that, though, I keep wishing all this SSD stuff stops being a platform feature. I'd love to see NVMe RAID and SSD caching become something supported by the relevant specifications and natively supported by OSes and their file systems rather than that just being an Intel/proprietary software thing (at the very least Windows needs to be better at storage heierarchy, even with caching losing relevance as fast and large SSDs go down in cost, the prevalence of cloud services and home servers has increased the need to better handle tertiary storage).
 

EdZ

Virtual Realist
May 11, 2015
1,578
2,107
I'd love to know if VROC also means Optane acceleration with the drives connected to the CPU (i.e. that RST can work through CPU lanes). It would be a real pain for X299 to be left out of Optane because everyone has connected the m.2 lanes to the CPU but the Optane caching hardware is still on the PCH. At least Optane Memory caching is listed explicitly, but IIRC no SODIMM Optane modules have been announced yet.
 

jØrd

S̳C̳S̳I̳ ̳f̳o̳r̳ ̳l̳i̳f̳e̳
sudocide.dev
SFFn Staff
Gold Supporter
LOSIAS
Jul 19, 2015
818
1,359
I'd love to see NVMe RAID and SSD caching become something supported by the relevant specifications and natively supported by OSes

Correct me if im wrong but hasnt this been a thing basically everywhere (which im defining as Linux & BSD for the moment) since forever. ZFS has fantastic caching functionality, LVM has Ok caching, RAM disks are a native thing on *nix & *nix based OS's, etc. Unless im missing something obvious (wouldnt be the first time) there isnt anything stopping me from picking up an Intel Optane SSD (or any other SSD) and using it as a cache drive basically anywhere i would care to except on Windows.
 

BirdofPrey

Standards Guru
Sep 3, 2015
797
493
Correct me if im wrong but hasnt this been a thing basically everywhere (which im defining as Linux & BSD for the moment) since forever. ZFS has fantastic caching functionality, LVM has Ok caching, RAM disks are a native thing on *nix & *nix based OS's, etc. Unless im missing something obvious (wouldnt be the first time) there isnt anything stopping me from picking up an Intel Optane SSD (or any other SSD) and using it as a cache drive basically anywhere i would care to except on Windows.
Well Intel is stopping you from using an Optane on anything they don't want you to.

As for caching, for the consumer space Windows pretty much IS "everywhere", Linux is only a tiny portion of users and I am pretty sure BSD is lumped in under "other" so that makes the OS-X implementation much more relevant. I do suppose this is more of a Windows rant, though, and I'd still like to see that as a hardware feature under NVMe and AHCI rather than being embedded into SSHDs or having the choice between Intel SRT and hoping maybe one day AMD will jump on the bandwagon.
Dealing with networked or cloud storage is a different matter. All of the major OSes out there deal reasonably well with being able to sync cloud/networked files, but it's also not as automatic as it could be sometimes.
 

EdZ

Virtual Realist
May 11, 2015
1,578
2,107
Well Intel is stopping you from using an Optane on anything they don't want you to.
You can use the Optane 'cache' SSDs anywhere you want to, they are exposed as an NVME m.2 drive just like any other NAND based NVME m.2 drive. You could throw one on a Ryzen board running Linux and set it up as the L2ARC drive for a ZFS filesystem with no problems.
The 'special sauce' is the same RST on-CPU/on-PCH caching (rather than in-software caching) that has been around for years, originally intended to add small m.2 NAND SATA SSD caches to spinning-rust drives. With Optane, this is extended to NVME SSDs also (and you can accelerate other NVME drives as well as SATA HDDs and SSDs). I think the only thing Intel is stopping anyone from doing is using an NVME NAND SSD (regardless of vendor) as a cache drive for another NVME NAND SSD.
 

EdZ

Virtual Realist
May 11, 2015
1,578
2,107
I was under the impression Intel was only allowing Optane to run on Kaby Lake and Later Intel systems
RST will only work to use an Optane drive as a transparent cache on Union Point (200-series) chipsets and onwards, but just using the Optane drive itself as a regular drive will work anywhere that supports NVME m.2 drives.
 
Mar 6, 2017
501
454
Details at PCWorld.com.

Apparently it is called Virtual RAID On CPU (VROC) and it only works on Skylake-X, not Kaby Lake-X.

If you're using Skylake-X, up to 20 storage devices can be set up into a virtual RAID 0.

If you want the array to be bootable, the storage devices all need to be Intel.

And then, if you want RAID 1, RAID 5, or other RAID options, you need to buy a special dongle with a key that enables "Premium Mode" in the BIOS.





So the answer is yes, but at a TBA price (enterprise feature = enterprise price).


Intel is becoming Apple...
 

QuantumBraced

Master of Cramming
Mar 9, 2017
507
358
Respect to ASRock, it doesn't get any better for an SFF enthusiast than getting the news about this board. My X99E-ITX/ac is my favorite piece of tech that I own (and one of my favorite things that I own, period) and a veritable engineering miracle. And this board seems to have outdone it in nearly every way.

I won't repeat all the praise that's already been given on the major improvements. A few smaller points - I am really glad they relocated the 24-pin. It was a pain to connect the included 24-pin cable of the Corsair SF600 to the X99E-ITX/ac in my NCase M1. It barely reached and had to stretch it diagonally across the board. The new placement will allow for much cleaner cable management. Also, I am really happy that they went with rigidly mounted Wi-Fi connectors. I hate the I/O shield-mounted connections on my X99E-ITX/ac, they get loose over time. So kudos on fixing that. And the board has one more power phase compared to the previous iteration -- not sure if that was really needed as 6-phase was already all you needed for maximum OC, but it can't hurt.

Question - Will this board come with an onboard headphone amp like the previous one? I use mine to drive my Sennheisers, saved me from having to buy an external DAC. I hope they kept this feature.

Question 2 - Does this board still come with 3 fan headers? Where are they? One major flaw of the previous board was that one of the headers was under the PCIe slot.

On aesthetics - I weirdly think it looks slightly worse. I liked the light blue theme of the previous iteration, it was a nice, soothing color. I actually made my system's entire color scheme light blue because of it. This one is very dark and Darth Vader-y. But I guess it's neutral which is good. It would be nice if they used a nicer font for all the text on the PCBs rather than this Arial or whatever it is, but I guess that's nitpicking. I also think there are pros and cons to having 4 SODIMM slots. Yes, you get quad-channel memory, but you have to deal with SODIMM's limitations. Unless you use very specific applications whose performance actually scales with memory bandwidth -- and I think that is a small minority of people -- it makes no difference. But I think overall, it's better to have the option.
 
  • Like
Reactions: WillBlackstone

royalba94

Airflow Optimizer
AVAENTUM SLEEVING
Apr 2, 2017
268
434
Avaentum.com
So will this board support Thunderbolt 3? I feel like I have heard it both ways...

Also, if not are there any other Mini-ITX mobos that support (or will support) Thunderbolt 3 yet?
 

Boil

SFF Guru
Nov 11, 2015
1,253
1,094
So will this board support Thunderbolt 3? I feel like I have heard it both ways...

Also, if not are there any other Mini-ITX mobos that support (or will support) Thunderbolt 3 yet?

I think the only mITX motherboard that currently supports TB3 is the ASRock Fatal1ty Z270 Gaming-ITX/ac...?
 
  • Like
Reactions: royalba94

Phryq

Cable-Tie Ninja
Nov 13, 2016
217
71
www.AlbertMcKay.com
I would like it to not have any audio output, as I have my own DAC, but of course, I can't expect them to cater to my unique needs :p

I'm just so happy for the NVMe M.2 and quad ram... I wonder if there'll be a RYZEN version.

And so this will accept Cores as well as XEONs, right? It's basically the ultimate motherboard but tiny. I want to make a 1.5L passively cooled super computer!
 
  • Like
Reactions: SeñorDonut

jeshikat

Jessica. Wayward SFF.n Founder
Silver Supporter
Feb 22, 2015
4,969
4,781
I haven't had time to research it but my understanding is X299 does not support Xeons.
 

Dyson Poindexter

If there's empty space, it's too big!
Jun 25, 2015
55
62
Only 5 USB Type-A connectors is a bummer to me, especially as there is ample room left on the back IO plate.

Although, one neat perk is that this board serves as a proof-of-concept for 4 DIMM slots on ITX. If they can do it with X299, they can do it with the "regular" sockets.
 
  • Like
Reactions: SeñorDonut

Boil

SFF Guru
Nov 11, 2015
1,253
1,094
Only 5 USB Type-A connectors is a bummer to me, especially as there is ample room left on the back IO plate.

Although, one neat perk is that this board serves as a proof-of-concept for 4 DIMM slots on ITX. If they can do it with X299, they can do it with the "regular" sockets.

ASRock already did it, with the server version of the X99 ITX board; you know, the one Linus put into a Dan A4-SFX...