Intel 750 U.2/AIC SSD RAID0 (X3) Report – 5GB/S & 750K IOPS

With the posting of two Intel 750 SSD Reviews, and our more recent article clarifying the U2 (SFF-8639) connector on the Intel 750, this NVMe SSD has received a bit of play as of late. Considering we have three 1.2TB 750s on hand, a few RAID tests were in order and we thought we might compile a bit of a report to get these numbers out.

Intel 750 NVMe SSDs with U2 Adapter

Two of our 750 Series NVMe SSDs are of the newest U2 (SFF-8639) form factor and the other is an AIC (add-in Card). Our thanks to ASRock for sending us two U.2 to M2 adapters, as these are still impossible to find in open retail. These adapters fit into the M.2 slot of newer motherboards and, as long as that connector accommodates PCIe 3.0 X4, loss of performance is no concern whatsoever.

ASRock M2 to U2 AdapterIn this scenario, we are using the ASRock Z170 Extreme7 motherboard and it houses three such M.2 slots. In addition, ASRock tackled all angles in setting up the UEFI with the latest version of Intel RST, enabling the RAIDing of all three M.2 SSDs right in the BIOS. We have tested this combination, as well as a few other SSD RAID scenarios, in both the BIOS RAID and OS software RAID setup and were able to confirm that performance in both is the same.

TSSDR Z170 Test Bench Intel 750 NVMe SSD RAID x3 Tests

This picture of our newest TSSDR Z170 Test Bench shows it hard at work testing the Intel 750 Series NVMe SSDs in RAID 0.


Our Z170 Test Bench was built with the sponsorship of Corsair, ASRock, Intel, and PNY, all of whom have our sincere appreciation. This is a relatively high-end PC and all of the components used for testing can be purchased at a reasonable price. The links below can assist in pricing and availability for those interested in this equipment:

PC CHASSIS: Corsair 760T White Full Window
CPU: Intel I7-6700K 4.00 GHz
CPU COOLER: Corsair Hydro Series H110i GTX
POWER SUPPLY: Corsair HX1200i  ATX12V
MEMORY: Corsair Dominator Pl  32GB 2800
 STORAGE: Intel 750 Series 1.2TB NVMe SSD
KEYBOARD: Corsair Vengeance K70 Mech
MOUSE: Corsair White M65 Laser
OS Microsoft Windows 10 Home 64 Bit


The software used for this report is typical of many of our reviews and consists of ATTO Disk Benchmark, Crystal DiskMark, AS SSD, Anvil Storage Utilities, PCMark Vantage and IOMeter. In consumer reports, we prefer to test with easily accessible software that the reader can obtain, and in many cases, we even provide the links. Our selection of software allows each to build on the last and to provide validation of results already obtained.


ATTO Disk Benchmark is a relatively easy-to-use benchmark tool, which happens to be the benchmark of choice for many manufacturers. ATTO uses compressible data rather than random data, which results in higher performance and thus, higher benchmark scores. In our testing, we have selected the transfer size to range from 0.5KB to 8192KB, and have set the total length of the test to be 256MB.

Intel 750 NVMe RAID X3 ATTO

We can see that ATTO wasn’t so comfortable testing in this particular NVMe RAID scenario which isn’t really much of a surprise.  We would like to have seen a steady speed increase commensurate with data sample size and high speeds of 4GB/s read and 3.7GB/s write weren’t quite what we were hoping to see.  Several tests were conducted to confirm this, all having similar results.

Leave a Reply

6 Comment threads
1 Thread replies
Most reacted comment
Hottest comment thread
7 Comment authors
Christoph WeihsdataherderMikeandydravo1 Recent comment authors
newest oldest most voted
Notify of
KB Grand
KB Grand

Just a great article! Really good job on the test. I thought the DMI 3 speed of 3.93 GB/s would affect performance, and yet it didn’t. How come?

Christoph Weihs
Christoph Weihs

because they used 1 pcie ssd which is connected to the cpu directly
the other 2 are connected through dmi

Lubomir Zvolensky
Lubomir Zvolensky

Guys, I believe: 1) [b]ATTO READ results are victim of 32bit overflow [/b]. 1024kB block show result of 4053109kB/s (=4GB/s), 2048kB block size only shows 700729kB/s caused by the 32bit overflow. This is AFTER it crossed 4GB/s threshold (2^32 = 4194304kB/s) so the TOTAL write speed in fact was 4GB/s + 0.7GB/s = 4.7GB/s. As I say, ATTO can NOT count with bigger than 32bit numbers internally, so that’s the reason why it only displays 0.7GB/s instead of what it should be 4.7GB/s. The same thing applies to 4096kB and 8192kB block size results, they SHOULD be 4.96GB/s and 4.99GB/s.… Read more »


We’re starting to hit the really intoxicating performance levels here. When you consider that you can get 2 800Gb units for about $800 and RAID0 them that’s a damn bargain.


It’s interesting how Intel can provide 3 new 750s for a raid 0 review, but can’t provide 1 mainstream 535 2.5 ssd for review to any site and it’s been out for a while now.