If you’ve heard a lot of buzz on solid state drives (SSDs) in recent months, you may be curious about what’s so special about them. After all, they do much the same thing as hard drives (HDDs) in terms of non volatile storage (i.e. boot up, and store applications and personal files). And HDDs also afford greater variety, availability and cost effectiveness. So what’s all the excitement about?
As it turns out, it’s not the result, but the mechanics of getting there that make a world of difference to your experience.
SSDs have brains (a controller that determines read/write operations), so they can greatly speed up your activity, reduce noise and heat, and consume less power. They can also be more resistant to physical damage. Since there is no rotating disc as in HDDs, they also occupy less space and can potentially be made denser in the future. Increasing adoption is also bringing down cost as manufacturers scale better, so SSD prices are on a decline.
The Story Behind The Story
But not all SSDs are made equal. Depending on the technology used, NAND memory comes in triple level cell (TLC), multi-level cell (MLC) and single level cell (SLC) structures. To cut a long story short, the NAND technology used has an effect on the SSD’s longevity, speed, reliability and price.
TLCs are the lightest on these attributes and are therefore the cheapest. MLCs are relatively more powerful and therefore more expensive and SLCs even more so. This means that consumer electronic devices that have shorter product life cycles can make do with TLCs while enterprise and data center operations require the more powerful varieties.
SSDs also come in various sizes: typical form factors mirror HHDs (5.25-inch, 3.5-inch, 2.5-inch or 1.8-inch). As you can imagine, the more storage-hungry operations require the larger sizes (for example, 5.25 inch used in backup devices). PCs and enterprise systems can usually make do with 3.5 inch SSDs. Most of the modern-day portables use 2.5 inch SSDs (or 1.8 inch in case of ultra-mobile), sacrificing storage capacity because of space constraints.
The Best of Both Worlds
In the initial stages, when adoption was lower and costs very high, some players including Samsung, Seagate and Western Digital thought they could pile a few SSDs onto the existing HDD to create what they called hybrids. This could speed up certain operations like booting and launching regularly-used apps while falling back on the HDD for other applications and storage. The increased complications of the technology however contributed to a further increase in cost.
Later, it was thought that a dual-drive system comprising a flash drive for the OS and apps and a hard drive for personal files would be a good idea. But estimating the amount of flash in this case was often a challenge and a dual drive system also needed more space.
Back to the Future
HDDs, SSDs and hybrids co-exist at present and may be expected to do so in the foreseeable future. That’s because the cost differential, density and production complexities are still being worked out. But the competitive landscape is shaping up as we speak.
Samsung Set to Reign Supreme
TLCs were traditionally used in flash memory cards, USB drives, client SSDs and cheaper smartphones. That was before Apple AAPL decided to use them in its iPhone 6. While there’s no actual information about Apple sourcing these from Samsung, the two appear really cozy at the moment with Samsung setting up a separate facility to accommodate Apple.
On the other hand, SanDisk SNDK, the other major TLC maker (and Apple supplier) reported the loss of a major SSD customer and significantly lowered guidance. So the street has been putting two and two together. If Samsung is really supplying Apple and consuming its own chips, it has cornered the lion’s share of the market. Samsung’s share of SSD units went from 33.3% in the fourth quarter of 2014 to 40.6% in the first quarter of 2015 according to the Storage Newsletter.
The Korean company is reportedly willing to sacrifice margins for market share. TLCs may grow faster than other categories although more players will also jump in.
SanDisk-Toshiba Missteps Could Prove Expensive
The duopolistic SSD market position that SanDisk once enjoyed with Samsung has been lost entirely. While Samsung remains extremely strong and ahead of the pack in terms of memory chip architecture, poor leadership, misreading of market trends, slow reaction to changing market conditions and other execution issues have plagued SanDisk until it finally lost important customers in both client and enterprise segments.
SanDisk is also handicapped in that it is a fabless pureplay SSD provider, which means that it is not in a position to profit from NAND sales, or drive production efficiencies and on the other hand has to transfer some profits to Toshiba, which fabricates its chips. Here too, it is exposed to the Yen. It does have a few good products in the pipeline that will hopefully make it to the market in 2016.
Whenever this actually happens, its joint investment in Toshiba’s Yokkaichi fab will come in handy (the fab is being pulled down and equipped to manufacture chips based on the latest 3D NAND technology). In the meantime though, prospective customers like HP (and maybe Qualcomm) will have long deflected to you guessed it, Samsung.
Intel INTC-Micron MUStrong Challenger
A major challenge for memory makers is that they cannot keep shrinking the size without compromising on the reliability of read/write operations. So just as in densely populated areas we have taller buildings to accommodate more people, memory chip structures are being changed to house more cells for storing more bits.
Simplistically put, manufacturers are targeting a U shape, with the base size being 1x nm (10 nm is thought to be the limit to which NAND can be shrunk) and each of the arms of the U accommodating 128 cells for a total of 256 cells. This structure is being referred to as 3D NAND because of its length, breadth and height. The increased storage space at the cost of producing this single and more capable chip is expected to make 3D NAND-based SSDs more competitive with HDDs.
In 2006, Intel and Micron formed the Intel Micron Flash Technology (IMFT) joint venture and in 2012, they amended the agreement according to which, Intel transferred some assets to Micron, which agreed to supply NAND chips to Intel. At its analyst day in Nov 2014, Intel said that the 3D NAND the JV produced was in prototype phase and would ship in 2H15.
Earlier this year, it announced a breakthrough product that could enable 10TB of storage, putting it in the lead as far as 3D NAND technology development is concerned. Intel has been the most consistent and transparent about its roadmap, so its projections appear most reliable.
This doesn’t of course mean that Intel and Micron will overthrow Samsung in the near future. Samsung is already shipping 3D NAND of lower density (roughly 4TB) most probably at a loss and likely has greater densities in the works. Intel-Micron on the other hand is yet to ship.
Intel technology also uses both MLC and TLC NAND (for enterprise and client SSDs) and is likely to be more expensive, since it will sell at a profit. Intel is also looking to further increase density by developing quad-level cell (QLC) technology, which is an industry-first, and so has a few skeptics that are concerned about reliability.
So What Happens to HDD Players Seagate STX and Western Digital WDC?
Since HDDs are not going away any time soon, these as well as hybrid devices will continue to generate revenue. But as SSDs eat into the core market, profitability will come under pressure. Not just that, the two players are trying to build their own SSD capabilities through internal development and acquisitions. This is easier said than done because increased investments are again pressures on profitability.
Moreover, the SSD market is very different from the HDD market in that it is far more competitive. Western Digital and Seagate own 90% of the HDD market, but they are late entrants into the SSD space where there are a number of strong and well-established players.
Since they have less time to innovate, the two players have gone the acquisition route and their common strategy is to protect enterprise market share first. This makes sense because initially SSDs will be expensive, so pickup on the consumer side can be slower. They will likely price aggressively.
Western Digital’s HGST picked up Skyera, which boosted its MLC capabilities. It also teamed up with Micron for product development.
Seagate is not to be outdone. The company acquired LSI’s Sandforce PCIe flash business from Avago AVGO. It currently sources flash from both Samsung and Toshiba.
Summing Up
Trendforce, as quoted by Forbes estimates that 3D NAND’s share of total NAND shipments will touch 20% this year and IHS estimates that this will jump to 65.7% in 2017. Since 3D NAND will speed up SSD adoption, these numbers are encouraging. Storage Newsletter estimates that enterprise SSDs can grow into a $4 billion market by 2016.
Samsung will likely get a ton of business on the client side and it’s still a bit early to tell if this will give it the scale to offset low prices and high costs of manufacturing SSDs (yields are likely very low given that process technology enabling this transition is still struggling to keep up). There will be a tooth and nail fight for the rest of the market.
The enterprise side is likely far more competitive from the get-go, with Samsung, SanDisk-Toshiba, Intel-Micron and soon, Seagate and Western Digital. Most players will have products based on the latest technology. But San Disk appears to have missed the bus; it is unlikely to have competitive technology until 2016-end or maybe even 2017.
Want the latest recommendations from Zacks Investment Research? Today, you can download 7 Best Stocks for the Next 30 Days. Click to get this free report >>
Want the latest recommendations from Zacks Investment Research? Today, you can download 7 Best Stocks for the Next 30 Days. Click to get this free report
Be the first to comment