Introduction, Specifications, and Requirements
Finally! Optane Memory sitting in our lab! Sure, it’s not the mighty P4800X we remotely tested over the past month, but this is right here, sitting on my desk. It’s shipping, too, meaning it could be sitting on your desk (or more importantly, in your PC) in just a matter of days.
The big deal about Optane is that it uses XPoint Memory, which has fast-as-lightning (faster, actually) response times of less than 10 microseconds. Compare this to the fastest modern NAND flash at ~90 microseconds, and the differences are going to add up fast. What’s wonderful about these response times is that they still hold true even when scaling an Optane product all the way down to just one or two dies of storage capacity. When you consider that managing fewer dies means less work for the controller, we can see latencies fall even further in some cases (as we will see later).
Introduction and Specifications
XPoint. Optane. QuantX. We've been hearing these terms thrown around for two years now. A form of 3D stackable non-volatile memory that promised 10x the density of DRAM and 1000x the speed and endurance of NAND. These were bold statements, and over the following months, we would see them misunderstood and misconstrued by many in the industry. These misconceptions were further amplified by some poor demo choices on the part of Intel (fortunately countered by some better choices made by Micron). Fortunately cooler heads prevailed as Jim Handy and other industry analysts helped explain that a 1000x improvement at the die level does not translate to the same improvement at the device level, especially when the first round of devices must comply with what will soon become a legacy method of connecting a persistent storage device to a PC.
Did I just suggest that PCIe 3.0 and the NVMe protocol - developed just for high-speed storage, is already legacy tech? Well, sorta.
That 'Future NVM' bar at the bottom of that chart there was a 2-year old prototype iteration of what is now Optane. Note that while NVMe was able to shrink down the yellow bar a bit, as you introduce faster and faster storage, the rest of the equation (meaning software, including the OS kernel) starts to have a larger and larger impact on limiting the ultimate speed of the device.
NAND Flash simplified schematic (via Wikipedia)
Before getting into the first retail product to push all of these links in the storage chain to the limit, let's explain how XPoint works and what makes it faster. Taking random writes as an example, NAND Flash (above) must program cells in pages and erase cells in blocks. As modern flash has increased in capacity, the sizes of those pages and blocks have scaled up roughly proportionally. At present day we are at pages >4KB and block sizes in the megabytes. When it comes to randomly writing to an already full section of flash, simply changing the contents of one byte on one page requires the clearing and rewriting of the entire block. The difference between what you wanted to write and what the flash had to rewrite to accomplish that operation is called the write amplification factor. It's something that must be dealt with when it comes to flash memory management, but for XPoint it is a completely different story:
XPoint is bit addressible. The 'cross' structure means you can select very small groups of data via Wordlines, with the ultimate selection resolving down to a single bit.
Since the programmed element effectively acts as a resistor, its output is read directly and quickly. Even better - none of that write amplification nonsense mentioned above applies here at all. There are no pages or blocks. If you want to write a byte, go ahead. Even better is that the bits can be changed regardless of their former state, meaning no erase or clear cycle must take place before writing - you just overwrite directly over what was previously stored. Is that 1000x faster / 1000x more write endurance than NAND thing starting to make more sense now?
Ok, with all of the background out of the way, let's get into the meat of the story. I present the P4800X:
ADATA has added another line of M.2 PCIe SSDs to their catalog with the XPG SX7000. These drives support NVMe and claim up to 1800 MB/s sequential read performance and 850 MB/s sequential write performance, with both tests measured on CrystalDiskMark at a queue depth of 32. Interestingly enough, their ATTO sequential write results, 860 MB/s, exceed their claimed maximum. Again, each of these numbers are provided by ADATA, so it’s still up to third-parties (like us) to verify. That said, ADATA provided a lot of information in their performance chart, which is nice to see.
The spec sheet (pdf) provides performance results for three SKUs: 128GB, 256GB, and 512GB. A fourth model (if you guessed 1TB, then you would be right) is also acknowledged, but not elaborated upon. These are all based on 3D TLC flash, with some undefined amount of SLC cache.
Pricing and availability are TBD, but it will come with a 5 year warranty.
Subject: Storage | April 7, 2017 - 07:01 AM | Scott Michaud
Tagged: WD, ssd, external ssd
Western Digital has just announced the My Passport SSD line of portable solid state hard drives. As you might expect, the major advantage of SSD-based portable storage is speed. This one connects with a USB Type-C port and is rated at up to 515 MB/s, although that hasn’t been benchmarked yet. The drives also support hardware, 256-bit AES encryption via their security software.
According to Best Buy, the 256GB model ($99.99 USD) is already sold out, but the 512GB model ($199.99) and the 1TB model ($399.99) are both still available for the 14th of April.
Subject: Storage | April 3, 2017 - 03:53 PM | Jeremy Hellstrom
Tagged: Noontec-TerraMaster, DAS, D2-310, usb 3.1
For those who don't want to get into networked storage solutions but still require external storage with more options than a simple USB drive offers, direct attached storage devices are a good solution. The Noontec-TerraMaster D2-310 is an aluminium shell with two drive bays, connected via Type-C USB 3.1 and offers support for JBOD, RAID 0 and RAID 1 in addition to simply presenting two external disks. Modders-Inc tested this DAS in two different configurations, a pair of Seagate 4 TB 7200 RPM HDDs as well as a pair of Samsung 850 EVO 256 SSDs. The performance levels reached their expectations, however the price is a bit higher than the competition; examine their results and description of the device to determine if you feel it is worth the expense.
"D2-310 is a direct attached storage device by Noontec-TerraMaster. Most of the market is moving away from DAS devices to network based devices however, there is still a need for simple and fast solutions to store data locally. D2-310 offers USB 3.1 connectivity and supports RAID redundancy in a two bay shell."
Here are some more Storage reviews from around the web:
- Buffalo MiniStation Velocity 960GB external SSD @ Kitguru
- ICY DOCK ICYCube Quad Bay 2.5" & 3.5" SATA External HDD Enclosure Review @ NikKTech
- Glyph 2TB AtomRAID Portable SSD @ The SSD Review
- WD Black PCIe NVMe @ The SSD Review
- Toshiba P300 3TB HDD @ Kitguru
- Intel gives hard drives a boost with Optane Memory @ The Tech Report
Introduction and Packaging
Data Robotics shipped their first product 10 years ago. Dubbed the Drobo (short for Data Robot), it was a 4-bay hot-swappable USB 2.0 connected external storage device. At a time where RAID was still a term mostly unknown to typical PC users, the Drobo was already pushing the concept of data redundancy past what those familiar with RAID were used to. BeyondRAID offered a form of redundant data storage that decoupled rigid RAID structures from fixed capacity disk packs. While most RAID volumes were 'dumb', BeyondRAID was aware of what was stored within its partitions, distributing that data in block format across the available disks. This not only significantly speed up rebuilding (only used portions of the disks need be recopied), it allowed for other cool tricks like the ability to mix drive capacities within the same array. Switching between parity levels could also be done on-the-fly and with significantly less effort than traditional RAID migrations.
While all of the above was great, the original Drobo saw performance hits from its block level management, which was limited by the processing overhead combined with the available processing power for such a device at the time. The first Drobo model was lucky to break 15 MB/s, which could not even fully saturate a USB 2.0 link. After the launch, requests for network attached capability led to the launch of the DroboShare, which could act as a USB to ethernet bridge. It worked but was still limited by the link speed of the connected Drobo. A Drobo FS launched a few years later, but it was not much quicker. Three years after that we got the 5N, which was finally a worthy contender in the space.
10 years and nearly a dozen models later, we now have the Drobo 5N2, which will replace the aging 5N. The newer model retains the same 5-bay form factor and mSATA bay for optional SSD cache but adds a second bondable Gigabit Ethernet port and upgrades most of the internals. Faster hardware specs and newer more capable firmware enables increased throughput and volume sizes up to 64TB. Since BeyondRAID is thin provisioned, you always make the volume as large as it can be and simply add disk capacity as the amount of stored content grows over time.
Today Samsung released an update to their EVO+ microSD card line. The new model is the 'EVO Plus'. Yes, I know, it's confusing to me as well, especially when trying to research the new vs. old iterations for this mini-review. Here's a few quick visual comparisons between both models:
On the left, we have the 'older' version of the Plus (I mean the '+'), while on the right we have the new plus, designated as a '2017 model' on the Samsung site. Note the rating differences between the two. The '+' on the left is rated at UHS-I U1 (10 MB/s minimum write speed), while the newer 'Plus' version is rated at UHS-I U3 (30 MB/s minimum write speed). I also ran across what looked like the older version packaging.
The packaging on the right is what we had in hand for this review. The image on the left was found at the Samsung website, and confuses things even further, as the 'Plus' on the package does not match the markings on the card itself ('+'). It looks as if Samsung may have silently updated the specs of the 256GB '+' model at some point in the recent past, as that model claims significantly faster write speeds (90 MB/s) than the older/other '+' models previously claimed (~20 MB/s). With that confusion out of the way, let's dig into the specs of this newest EVO Plus:
For clarification on the Speed Class and Grade, I direct you to our previous article covering those aspects in detail. For here I'll briefly state that the interface can handle 104 MB/s while the media itself is required to sustain a minimum of 30 MB/s of typical streaming recorded content. The specs go on to claim 100MB/s reads and 90 MB/s writes (60 MB/s for the 64GB model). Doing some quick checks, here's what I saw with some simple file copies to and from a 128GB EVO Plus:
Our figures didn't exceed the specified performance, but they came close, which more than satisfies their 'up to' claim, with over 80 MB/s writes and 93 MB/s reads. I was able to separately confirm 85-89 MB/s writes and 99 MB/s reads with Iometer accessing with 128KB sequential transfers.
- 32GB: $29.99
- 64GB: $49.99
- 128GB: $99.99
- 256GB: coming soon (but there is already a 256GB EVO+ of similar specs???)
Pricing seems to be running a bit high on these, with pricing running close to double of the previous version of this very same part (the EVO+ 128GB can be found for $50 at the time of this writing). Sure you are getting a U3 rated card with over four times the achievable write speed, but the reads are very similar, and if your camera only requires U1 speeds, the price premium does not seem to be worthwhile. It is also worth noting that even faster UHS-II spec cards that transfer at 150 MB/s can be had and even come with a reader at a lower cost.
In summary, the Samsung EVO Plus microSD cards look to be decent performers, but the pricing needs to come down some to be truly competitive in this space. I'd also like to see the product labeling and marketing a bit more clear between the '+' and the 'Plus' models, as they can easily confuse those not so familiar with SD card classes and grades. It also makes searching for them rather difficult, as most search engines parse 'Plus' interchangeably with '+', adding to the potential confusion.
The Need for Speed
Around here storage is Allyn’s territory, but I decided to share my experience with a new $20 flash drive I picked up that promised some impressive speeds via USB 3.0. The drive is the Lexar JumpDrive P20, and I bought the 32GB version, which is the lowest capacity of the three drives in the series. 64GB and 128GB versions of the JumpDrive P20 are available, with advertised speeds of up to 400 MB/s from all three, and reads and up to 270 MB/s writes - if you buy the largest capacity.
My humble 32GB model still boasts up to 140 MB/s writes, which would be faster than any USB drive I’ve ever owned (my SanDisk Extreme USB 3.0 16GB drive is limited to 60 MB/s writes, and can hit about 190 MB/s reads), and the speeds of the P20 even approach that of some lower capacity SATA 3 SSDs - if it lives up to the claims. The price was right, so I took the plunge. (My hard-earned $20 at stake!)
Size comparison with other USB flash drives on hand (P20 on far right)
First we'll look at the features from Lexar:
- Among the fastest USB flash drives available, with speeds up to 400MB/s read and 270MB/s write
- Sleek design with metal alloy base and high-gloss mirror finish top
- Securely protects files using EncryptStick Lite software, an advanced security solution with 256-bit AES encryption
- Reliably stores and transfers files, photos, videos, and more
- High-capacity options to store more files on the go
- Compatible with PC and Mac systems
- Backwards compatible with USB 2.0 devices
- Limited lifetime warranty
Subject: Storage | March 27, 2017 - 12:16 PM | Allyn Malventano
Tagged: XPoint, Optane Memory, Optane, M.2, Intel, cache, 3D XPoint
We are just about to hit two years since Intel and Micron jointly launched 3D XPoint, and there have certainly been a lot of stories about it since. Intel officially launched the P4800X last week, and this week they are officially launching Optane Memory. The base level information about Optane Memory is mostly unchanged, however, we do have a slide deck we are allowed to pick from to point out some of the things we can look forward to once the new tech starts hitting devices you can own.
Alright, so this is Optane Memory in a nutshell. Put some XPoint memory on an M.2 form factor device, leverage Intel's SRT caching tech, and you get a 16GB or 32GB cache laid over your system's primary HDD.
To help explain what good Optane can do for typical desktop workloads, first we need to dig into Queue Depths a bit. Above are some examples of the typical QD various desktop applications run at. This data is from direct IO trace captures of systems in actual use. Now that we've established that the majority of desktop workloads operate at very low Queue Depths (<= 4), lets see where Optane performance falls relative to other storage technologies:
There's a bit to digest in this chart, but let me walk you through it. The ranges tapering off show the percentage of IOs falling at the various Queue Depths, while the green, red, and orange lines ramping up to higher IOPS (right axis) show relative SSD performance at those same Queue Depths. The key to Optane's performance benefit here is that it can ramp up to full performance at very low QD's, while the other NAND-based parts require significantly higher parallel requests to achieve full rated performance. This is what will ultimately lead to a much snappier responsiveness for, well, just about anything hitting the storage. Fun fact - there is actually a HDD on that chart. It's the yellow line that you might have mistook as the horizontal axis :).
As you can see, we have a few integrators on board already. Official support requires a 270 series motherboard and Kaby Lake CPU, but it is possible that motherboard makers could backport the required NVMe v1.1 and Intel RST 15.5 requirements into older systems.
For those curious, if caching is the only way power users will be able to go with Optane, that's not the case. Atop that pyramid there sits an 'Intel Optane SSD', which should basically be a consumer version of the P4800X. It is sure to be an incredibly fast SSD, but that performance will most definitely come at a price!
We should be testing Optane Memory shortly and will finally have some publishable results of this new tech as soon as we can!
Subject: Storage | March 25, 2017 - 02:13 AM | Scott Michaud
Tagged: Lexar, thumb drive
A new line of USB flash drives has been announced by Lexar, which focuses on both durability and USB 3.1 support (compatible with USB 2.0 and USB 3.0). From the technical side, the Lexar JumpDrive Tough drives can read up to 150 MB/s and write up to 60 MB/s, which is obviously nowhere near SSD speed, but reasonably fast for the typical cases that you would use a thumb drive.
As for its robustness, Lexar claims that the JumpDrive Tough will operate normally between -13F and 300F, which is just shy of the bake cookies temperature. It is also water resistant up to 98 feet.
The Lexar JumpDrive Tough will be available in 32GB, 64GB, and 128GB models for $19.99, $34.99, and $59.99, respectively. While I don’t normally consider manufacturer returns for something like this, Lexar is backing this purchase with a 3-year limited warranty, which gives some legal teeth to their claims (if anyone takes them up on it). They are available now.