Introduction, Specifications and Packaging
A while back, we saw OCZ undergo a major restructuring. 150+ product SKUs were removed from their lineup, leaving a solid core group of products for the company to focus on. The Vertex and Agility lines were spared, and the Vector was introduced and well received by the community. With all of that product trimming, we were bound to see another release at some point:
Today we see a branch from one of those tree limbs in the form of the Vertex 3.20. This is basically a Vertex 3, but with the 25nm IMFT Sync flash replaced by newer 20nm IMFT Sync flash. The drop to 20nm comes with a slight penalty in write endurance (3000 cycles, down from the 5000 rating of 25nm) for the gain of cheaper production cost (more dies per 300mm wafer).
IMFT has been cooking up 20nm flash for a while now, and it is becoming mature enough to enter the mainstream. The first entrant was Intel's own 335 Series, which debuted late last year. 20nm flash has no real groundbreaking improvements other than the reduced size, so the hope is that this shrink will translate to lower cost/GB to the end user. Let's see how the new Vertex shakes out.
- Capacity: 120, 240GB
- Sequential read: 550 MB/sec
- Sequential write: 520 MB/sec
- Random read IOPS (up to): 35 k-IOPS
- Random write IOPS (up to): 65 k-IOPS
This simple plastic packaging does away with the 3.5" bracket previously included with all OCZ models.
Subject: Storage | March 28, 2013 - 04:15 PM | Jeremy Hellstrom
Tagged: SuperSSpeed, S301 Hyper Gold, ssd, slc, SandForce SF-2281
SuperSSpeed is mixing the performance and endurance of SLC flash storage with the lower cost of the SandForce SF-2281 in an attempt to bring the price of their SLC drive to an affordable level for the consumer. The mix seems a good idea as the reduced write latency of SLC flash may help to overcome SandForce's weakness when writing incompressible data. [H]ard|OCP's testing bears this out as the drive kept up with a larger Samsung 840 Pro, one of the current performance kings. You will pay for the privilege however as the 128GB drive currently retails for $250 as SLC flash is not cheap. Consider that in almost any casual usage scenario, you are never going to push this drive to its limits ... unless you are going to start your own Frame Rating machine.
"The SuperSSpeed S301 128GB SLC SSD brings SLC flash into the consumer market. The extreme endurance and excellent write performance makes for an interesting SSD powered by the SandForce SF-2281 controller. The Intel 25nm SLC NAND removes much of the Achilles heel of the SandForce processors, delivering consistent performance."
Here are some more Storage reviews from around the web:
- OCZ Vertex 3.20 – Vertex 3 updated to 20nm @ Bjorn3D
- OCZ Vertex 3.20 120GB Solid State Drive Review @ Pro-Clockers
- OCZ Vertex 3 .20 120GB SSD @ Tweaktown
- OCZ Vertex 3 .20 240GB SSD @ Tweaktown
- KingFast Ultra-Cache K13 & K25 SATA2 SSD Review @ ModSynergy
- Kingston V300 120GB SSD Review @ HCW
- Kingston V300 120GB SSD @ Bjorn3D
- Intel 335 Series 180GB SSD Review @ Hardware Canucks
- The SSD Review SSD Database Is Live
- ADATA DashDrive Air AE400 Wireless Storage Device @ Tweaktown
- Kingston DataTraveller Ultimate 3.0 G3 64GB USB3.0 Flash Drive @ Tweaktown
- Kingston HyperX Predator 512GB USB 3.0 Flash Drive @ Tweaktown
- QNAP TS-469L @ Legion Hardware
- G-Technology G-Drive Mobile USB 1TB USB 3.0 Portable Hard Drive Review @ NikKTech
- OWC Mercury On-The-Go Pro USB 3.0 Portable Enclosure Kit Review @ Madshrimps
- ADATA DashDrive HV610 External Hard Drive @ Tweaktown
- ADATA DashDrive Durable HD710 External Hard Drive @ Tweaktown
Subject: General Tech | March 20, 2013 - 01:02 PM | Jeremy Hellstrom
Tagged: DRAM, micron, ssd, Samsung, Hynix
It is perhaps not obvious to many because of the huge number of DRAM resellers but there are only three major manufacturers of DRAM left at this point. Apart from Micron, who claim top spot in this article on The Register, Samsung and Hynix are the only other big players left supplying DRAM. Considering the instability of memory and SSD pricing it seems odd that it is a component with only three possible sources, the instability could be coming from the fact that many of the mergers are still rather recent or in the case of Elpida, not quite complete yet. One very interesting comment from Kipp Bedard, Micron's investor relations VP, might also explain the volatilty of flash, "there simply isn't enough NAND fab capacity to store even 20 per cent of the data people are generating." If demand outstrips supply by that order of magnitude you can dictate almost any price you wish.
"When I first started at Micron, there were about 40 to 50 DRAM companies in the space," said Bedard. "And we spent most of the '80s with the Japanese deciding they wanted to own the DRAM space which they went from 10 per cent market share to about 90 per cent, [and] took all of the US companies out except for two, us and Texas Instruments."
Here is some more Tech News from around the web:
- Fusion-io gobbles Brit Linux SCSI gurus ID7 @ The Register
- Report: BlackBerry BYOD-ware doesn't pass UK.gov security test @ The Register
- Netatmo review: weather station with app @ Hardware.info
Subject: Storage | March 18, 2013 - 04:23 PM | Jeremy Hellstrom
Tagged: Solidata, K8 1920E 2TB, ssd, sf-1222, LSI, sandforce, Micron JMB393
We have seen some high capacity PCIe based SSDs but in the 2.5" form factor they have been few and far between. This will soon change as Solidata will be releasing a 2 Terabyte SSD called the K8 1920E which will be somewhere in the neighbourhood of $5000 when it becomes available. Each one of the flash storage chips you can see below is a 64GB chip and with 16 on each side you get a full 2048GB of storage. It uses four of the LSI Sandforce SF-1222 controllers and a Micron JMB393 SATA II RAID-5 controller which is configured to act as a 4 port hub, treating each of the controllers as a separate 512GB SSD. Once the SSD Review had formatted the drive for use there was a total of 1788GB available for storage which did not support TRIM as it is technically behind a RAID card. The performance was on par with expectations, keeping in mind the difficulties that SandForce controllers have with incompressible data. This drive will be very expensive but it seems it will be the first product of its type available to be purchased.
"Ever since SSDs were introduced to the retail market back in 07, one of the main complaints has always been capacity. After all, the first SSD releases were only 32 and 64GB. The hopes of one day seeing the performance of an SSD coupled with the capacity of a hard drive has grown and, too many, we think our analysis of the new Solidata K8-1920E 2TB SSD might be welcome news."
Here are some more Storage reviews from around the web:
- SuperSSpeed S301 Hyper Gold 128GB SLC SSD @ [H]ard|OCP
- Intel 335 Series 180GB SSD Review @ Techgage
- MyDigitalSSD BP4 Slim 7 Solid State Drive @ Benchmark Reviews
- OCZ Vertex 3.20 20nm @ SSD Review
- Micron RealSSD P400m Enterprise SSD @ SSD Review
- Kingston SSDNow V300 120GB @ Tweaktown
- OCZ Vector 256GB SSD Review @ Hi Tech Legion
- Samsung 840 Pro 512GB @ Tweaktown
- Toshiba MK01GRRB/R 2.5-inch 6Gb/s SAS 15K RPM Enterprise RAID Report @ Tweaktown
- Patriot Gauntlet Node Wireless Enclosure Review @ NikKTech
- Adata DashDrive Air AE400 review: wireless card reader for mobile devices @ Hardware.info
- PQI Tiffany USB 3.0 32 GB @ techPowerUp
- Transcend RDF8 USB 3.0 Memory Card Reader Review @ Legit Reviews
- SuperTalent RC4 USB 3.0 Flash Drive With MS Windows To Go @ SSD Review
- Patriot Supersonic Magnum 256GB USB 3.0 Flash Drive Review @ Legit Reviews
- Patriot Supersonic Magnum 256GB USB 3.0 Flash Drive Review @ Hi Tech Legion
- CalDigit AV Pro USB 3.0 HDD / SSD Enclosure @ Tweaktown
- Thecus N7510 7-Bay Affordable Tower NAS @ Tweaktown
- QNAP TS-469L High-performance 4-bay NAS Server for Home & SOHO Review @ Madshrimps
- StarTech 2.5-Inch to USB 3.0 Encrypted Hard Drive Enclosure Review @ Legit Reviews
- QNAP TurboNAS TS-469U-RP NAS Server Review @ NikKTech
- Icy Dock FlexCage MB973SP-2B 5.25-inch HDD Bay Adapter @ Tweaktown
Apacer recently launched two new SSDs aimed at commercial and industrial applications. The drives will offer up either SLC or MLC NAND flash, but with a twist. The two drives feature the IDE / PATA interface instead of the newer SATA interface seen in today’s systems. Apacer is hoping its PATA SSDs will be used as an upgrade path when the hard drives currently used in industrial systems need replaced. The new Solid State Drives fall under the Apacer AFD 257 and AFD 187 series. The Apacer AFD-257 Premium is a 2.5" drive, and the AFD-187 Premium is a 1.8" drive.
To accommodate the greater need to data protection in such systems, Apacer has built several security features into the drives. The new PATA SSDs include Full Erase, Destroy, and Write Protection features. Interestingly, those security features can be activated using software or via hardware connected to a small port on the drives via a cable that can be routed to a control panel on the external IO of a chassis.
The drives have up to 256 GB capacities and have standard features such as SMART, wear leveling, and ECC (72-bit). The IDE interface is rather antiquated, but Apacer at least supports the faster transfer modes including: DIO Mode-4, MWDMA Mode-2, and Ultra DMA-6. PATA SSDs were somewhat-rare when IDE was still the dominant consumer standard, so it is nice to see there are alternatives for replacement parts still available.
Unfortunately, there is no word on pricing or availability. Transfer speeds are also unknown, but you can expect it to be bottle-necked by the IDE interface (though random access speeds should be a huge improvement over a hard drive, even with the slower PATA interface).
Subject: General Tech | February 27, 2013 - 01:11 PM | Jeremy Hellstrom
Tagged: ssd, enterprise ssd, SAS, micron, micron p410m
Micron has announced a new SSD, the P410 SSD which will use a Serial Attached SCSI interface, perfect for dropping into existing enterprise servers. SATA is perfectly fine for SOHO users and enthusiasts but for large businesses with a need for extreme reliability, SAS has been the interface of choice. Adoption of SSDs has been slowed in large businesses in part because it would require changing the existing architecture to SATA in order to incorporate SSDs into their systems. With the new Micron drive that is no longer necessary, at 7mm it will support high density servers and with the 25nm MLC NAND it is expected to survive for five years of duty with 10 full drive fills every day. Read more at DigiTimes.
"Micron Technology has announced another addition to its growing lineup of solid state drives (SSDs) targeted at data center appliances and enterprise storage platforms. The new Micron P410m SSD is a high-endurance, high reliability 6Gb/s Serial Attached SCSI (SAS) drive built to provide the performance necessary for mission-critical tier one storage applications that require uninterrupted, 24/7 data access."
Here is some more Tech News from around the web:
- AMD to sell a cut down version of Sony's Playstation 4 APU @ The Inquirer
- Google offers a single sign-on system, embraces 10 partners @ The Inquirer
- Benchmarking Ubuntu Linux On The Google Nexus 10 @ Phoronix
- Intel takes on all Hadoop disties to rule big data munching @ The Register
- Stuxnet worm dates back to 2005, Symantec reveals @ The Inquirer
- First Debian/Ubuntu Bootable ARM64 Images Released @ Slashdot
Subject: Storage, Shows and Expos | February 25, 2013 - 02:12 PM | Jeremy Hellstrom
Tagged: ocz, ssd, PCIe SSD, CeBIT 2013, ZD-XL Accelerator
SAN JOSE, CA—February 25, 2013—OCZ Technology Group, Inc. (Nasdaq:OCZ), a leading provider of high-performance solid-state drives (SSDs) for computing devices and systems, today announced that it will preview a variety of enterprise storage solutions at next week’s CeBIT 2013 conference in Hannover, Germany. As a renowned global forum, CeBIT represents a great opportunity for attendees to be the first to see and experience the latest innovations in solid-state storage from an industry leader in enterprise SSDs, virtualization, and caching software. OCZ offers a complete suite of storage solutions that address VMware, Linux, and SQL Server platforms, and invites IT decision-makers who are evaluating or implementing solid-state storage in the data center to visit the Company’s exhibit in Hall 2, Stand E43, from March 5th through 9th.
OCZ will unveil the next-generation ZD-XL SQL Accelerator, a culmination of enterprise hardware and software converging as one tightly integrated and optimized solution. The ZD-XL Accelerator addresses SQL Server database applications to not only ensure that the data for this implementation is right, relevant, and readily available on SSD flash when the SQL Server needs it, but also that the data is accessed with the highest possible I/O performance. For simple deployment and ease of use, this tightly integrated, optimized solution features ‘implementation wizards’ to guide DBAs so they can optimally manage data cached to the flash. While showcasing the ZD-XL solution OCZ will invite enterprise customers to become beta testers for this exciting solution.
Also included in OCZ’s exhibition at CeBIT will be demonstrations to preview the upcoming VXL 1.3 Virtualization Software and LXL Acceleration Software with OCZ’s innovative Direct Pass Caching Technology, which not only addresses VMware but is also optimized for Linux applications. OCZ is one of the few SSD providers with a robust portfolio of virtualization and caching software that combine the power of flash acceleration with the power of storage virtualization. This enables multiple virtual server loads to run concurrently on a single physical host, not only increasing CPU and memory resource utilization, but also simplifying deployment, high availability (HA), and maintenance of the server loads.
The next-generation of workstation PCI Express (PCIe)-based SSDs will also be available soon as part of the Company’s award-winning Vector Series. These drives reside directly on the PCIe bus and will support four PCIe Gen2 lanes providing lower latency to data, faster file transfers and system boot-ups, expanded storage capacities, and an even quicker, more responsive experience over the already blazingly fast SATA 3.0-based Vector Series. The Vector PCIe Series will feature an advanced suite of flash management tools that deliver enhanced drive endurance and data, making it ideally suited for power computing, content creation, and workstation applications.
Subject: Storage | February 22, 2013 - 01:33 PM | Jeremy Hellstrom
Tagged: Samsung 840, Samsung, ssd, 120gb, Samsung MDX
[H]ard|OCP just wrapped up a review of the 120GB Samsung 840, using their own ARM Cortex R4 based MDX controller and TLC memory for storage. They compare the speed of this drive to the 256GB 840 Pro, Kingston's V300 120GB and the Intel 335 240GB to contrast the difference the type of NAND used can make to performance. This is especially evident on the write and latency benchmarks, which fall well behind the competition. From [H]'s testing it is apparent that TLC memory is very vulnerable to reduction in size, the reduced channels really hurt performance and put the 120GB model far behind the larger sized 840s which they have tested with much better results.
"The 120GB Samsung 840 Series SSD features the powerful 8-channel MDX controller and TLC NAND. While this value SSD comes at a very good price, it also features much lower speeds than its larger capacity brethren. We put this value SSD through our suite of steady state tests to see if it can pass muster."
Here are some more Storage reviews from around the web:
- Mushkin Atlas 480GB SSD Review - Performance Meets Capacity In The Ultra World @ SSD Review
- MyDigitalSSD BP4 Slim 7 Series SSD @ SSD Review
- Corsair Neutron GTX 240GB SSD Review @ Techgage
- OCZ Vector 256GB Solid State Drive Review @ Pro-Clockers
- Kingfast F3 Plus SSD @ SSD Review
- TRIM Check: Overview of an essential SSD TRIM functionality tester @ Tweaktown
- Seagate Constellation ES.3 4TB Enterprise Hard Drive Review @ Techgage
- Two 2 Bay NAS Review: Synology DiskStation DS213+ and QNAP TS-269 Pro @ Custom PC Review
- Vantec NexStar NST-D300WS3 WiFi HDD Dock Review @ Hi Tech Legion
- LaCie 5big Thunderbolt: super-fast 20TB NAS device @ Hardware.info
- ASUSTOR AS-604T 4-bay NAS Server for Home and Small Business Review @ Madshrimps
- Asustor AS-604T NAS review: Worthy new competitor for QNAP and Synology @ Hardware.info
- ADATA DashDrive Elite 500GB - External USB3.0 Hard Drive @ FunkyKit
- Icy Dock Blizzard HDD Enclosure & EZ-Dock Docking Station @ Silent PC Review
- Patriot SuperSonic Magnum 256GB USB 3.0 Drive @ Kitguru
- ADATA DashDrive Air AE400 Wireless Storage Reader Review @ Custom PC Review
- Patriot Gauntlet 320GB Wireless Hard Drive @ eTeknix
Subject: General Tech | February 21, 2013 - 02:58 AM | Ken Addison
Tagged: titan, Tegra 4i, tegra 4, ssd, ps4, podcast, nvidia, Intel
PC Perspective Podcast #239 - 02/21/2013
Join us this week as we discuss NVIDIA GTX TITAN, PlayStation 4 Hardware, SSD Endurance and more!
The URL for the podcast is: http://pcper.com/podcast - Share with your friends!
- iTunes - Subscribe to the podcast directly through the Store
- RSS - Subscribe through your regular RSS reader
- MP3 - Direct download link to the MP3 file
Hosts: Ryan Shrout, Jeremy Hellstrom, Josh Walrath and Allyn Malventano
Program length: 0:59:53
Podcast topics of discussion:
- Week in Reviews:
- 0:27:20 This Podcast is brought to you by MSI!
- News items of interest:
0:49:00 Hardware / Software Pick of the Week
- Ryan: Excel 2013.. or not.
- Jeremy: WobbleWorks $75 Pen-Sized 3D Printer on Kickstarter
- Josh: Sweet lookin Monitor... not without quirks
- Allyn: Gunnar
- 0:49:00 Hardware / Software Pick of the Week
- 1-888-38-PCPER or firstname.lastname@example.org
- http://twitter.com/ryanshrout and http://twitter.com/pcper
Taking an Accurate Look at SSD Write Endurance
Last year, I posted a rebuttal to a paper describing the future of flash memory as ‘bleak’. The paper went through great (and convoluted) lengths to paint a tragic picture of flash memory endurance moving forward. Yesterday a newer paper hit Slashdot – this one doing just the opposite, and going as far as to assume production flash memory handling up to 1 Million erase cycles. You’d think that since I’m constantly pushing flash memory as a viable, reliable, and super-fast successor to Hard Disks (aka 'Spinning Rust'), that I’d just sit back on this one and let it fly. After all, it helps make my argument! Well, I can’t, because if there are errors published on a topic so important to me, it’s in the interest of journalistic integrity that I must now post an equal and opposite rebuttal to this one – even if it works against my case.
First I’m going to invite you to read through the paper in question. After doing so, I’m now going to pick it apart. Unfortunately I’m crunched for time today, so I’m going to reduce my dissertation into the form of some simple bulleted points:
- Max data write speed did not take into account 8/10 encoding, meaning 6Gb/sec = 600MB/sec, not 750MB/sec.
- The flash *page* size (8KB) and block sizes (2MB) chosen more closely resemble that of MLC parts (not SLC – see below for why this is important).
- The paper makes no reference to Write Amplification.
Perhaps the most glaring and significant is that all of the formulas, while correct, fail to consider the most important factor when dealing with flash memory writes – Write Amplification.
Before geting into it, I'll reference the excellent graphic that Anand put in his SSD Relapse piece:
SSD controllers combine smaller writes into larger ones in an attempt to speed up the effective write speed. This falls flat once all flash blocks have been written to at least once. From that point forward, the SSD must play musical chairs with the data on each and every small write. In a bad case, a single 4KB write turns into a 2MB write. For that example, Write Amplification would be a factor of 500, meaning the flash memory is cycled at 500x the rate calculated in the paper. Sure that’s an extreme example, but the point is that without referencing amplification at all, it is assumed to be a factor of 1, which would only be the case if you were only writing 2MB blocks of data to the SSD. This is almost never the case, regardless of Operating System.
After posters on Slashdot called out the author on his assumptions of rated P/E cycles, he went back and added two links to justify his figures. The problem is that the first links to a 2005 data sheet for 90nm SLC flash. Samsung’s 90nm flash was 1Gb per die (128MB). The packages were available with up to 4 dies each, and scaling up to a typical 16-chip SSD, that only gives you an 8GB SSD. Not very practical. That’s not to say 100k is an inaccurate figure for SLC endurance. It’s just a really bad reference to use is all. Here's a better one from the Flash Memory Summit a couple of years back:
The second link was a 2008 PR blast from Micron, based on their proposed pushing of the 34nm process to its limits. “One Million Write Cycles” was nothing more than a tag line for an achievement accomplished in a lab under ideal conditions. That figure was never reached in anything you could actually buy in a SATA SSD. A better reference would be from that same presentation at the Summit:
This shows larger process nodes hitting even beyond 1 million cycles (given sufficient additional error bits used for error correction), but remember it has to be something that is available and in a usable capacity to be practical for real world use, and that’s just not the case for the flash in the above chart.
At the end of the day, manufacturers must balance cost, capacity, and longevity. This forces a push towards smaller processes (for more capacity per cost), with the limit being how much endurance they are willing to give up in the process. In the end they choose based on what the customer needs. Enterprise use leans towards SLC or eMLC, as they are willing to spend more for the gain in endurance. Typical PC users get standard MLC and now even TLC, which are *good enough* for that application. It's worth noting that most SSD failures are not due to burning out all of the available flash P/E cycles. The vast majority are due to infant mortality failures of the controller or even due to buggy firmware. I've never written enough to any single consumer SSD (in normal operation) to wear out all of the flash. The closest I've come to a flash-related failure was when I had an ioDrive fail during testing by excessive heat causing a solder pad to lift on one of the flash chips.
All of this said, I’d love to see a revisit to the author’s well-structured paper – only based on the corrected assumptions I’ve outlined above. *That* is the type of paper I would reference when attempting to make *accurate* arguments for SSD endurance.