Given all the confusion surrounding Seagate’s stance on Solid State Disk technology, we decided to sit down with Seagate’s representatives with an internal knowledge of the subject in hand. We talked with David Szabados and Rich Vignes on where Seagate is headed with their Solid-State Disk Drive offerings. It was an enlightening discussion and filled in some missing information on why Seagate has been so quiet in the booming market of Solid-State Drives.
I started off by asking the obvious question whether Seagate thought it has missed the boat on the solid-state market. Rich replied by saying that Seagate wanted to avoid a “Me-Too” entry into the SSD playing field. Instead of just shoving another product on the consumer or enterprise segment, Seagate wanted to be different from the other manufacturers releasing SSDs today. As Seagate is aiming at the enterprise they needed to find out exactly what they wanted and to be sure that SSDs would meet the enterprise demands for robustness and performance. Rich assured us that not being first out of the door would not hurt Seagate’s products in the market.
Rich followed up by saying that Seagate wants to focus on the enterprise market for now and give them a product that will fit in with the current levels of performance and endurance expected in that space. Seagate has taken input from consumers to create a product to meet that demand. To start out they are going to begin with SLC-based products [due to customer input] but feel that the SSD must move to MLC [Multi-Level Cell] in order to be a viable volume product. The issue is that in its current form, MLC is not viable as an enterprise standard – only when this format becomes perfected you can expect Seagate to move to MLC.
But the decision to use SLC or MLC is not the only one Seagate is facing, they are also working to determine the appropriate connection to use; SAS or SATA-II. Each one has pros and cons to them. As it stands right now both are limited to a 3Gb/s transfer rate which in many cases can be quickly saturated by an SSDs or array of SSDs. This limitation that they share should be lifted soon as they both move to 6Gb/s in the near future – as you might already know, AMD is going to be the first company with SATA 3.0 interface, coming in the form of SB850 Southbridge chip. SAS [Serial Attached SCSI – Small Computer System Interface] has a couple of major advantages over SATA though. The two biggest are dual port and error correction. SAS has a much deeper level of error correction and can work all the way down to the media and communicate back to the host. SATA has some basic “best-effort” error correction but is not enough for business critical or mission critical applications. In the end I think that SAS will be the ultimate choice for SSDs in the enterprise segment. SAS offers better performance and robustness than what SATA can provide.
The next challenge for Seagate is industry acceptance of capacity; while we see consumer products with capacities as small as 32GB in an enterprise situation this might not be a usable capacity for anything other than a boot drive or for a high-performance swap file. Also, these smaller capacity drives can lose performance as the number of internal parts increase. This puts the 128 and 256GB drives as the main area of interest. Still there is no word on which capacity will eventually be selected. It will come down to a question of capacity Vs performance Vs price. Seagate could put out a 1TB SSD tomorrow but more than likely the cost of this drive would prevent its acceptance. If I were a betting man I would think that 256GB will be the entry point and then a 512GB will pop up within a year of launch.
Continued on the next page…
On the subject of heat Rich says that while on paper SSDs usually use less power and generate less heat, the overall power usage and heat consideration will vary from design to design. There are factors that make it hard to predict the max operating temperatures that an SSD can operate in as the smaller the individual parts the more they are affected by heat.
I asked Rich about Seagate’s plans for consumer drives and he says that while there is no set timeline he was sure that consumer versions of the enterprise SSDs would show up in the future and for everyone to keep an eye out for them. When I asked if Seagate would be looking into other flash products [such as USB keys etc] Rich let me know that Seagate has no interest in them whatsoever. It looks like this market is completely saturated with low-margin companies from the Far East and there is not a lot of innovation that Seagate could put in order to differentiate itself from the competition.
The new SSDs will all start off in a 2.5-inch form factor and will continue to use much of Seagate’s own IP as far as drive controllers. This is a very economical way of building a new product, after all Seagate has a large vault of IP to choose from when it comes to drive controllers and the applications are close enough that current technology should fit in nicely. As with most products there will be third-party parts but for the most part Seagate will be making its own controllers for their SSDs. For the Blade guys out there as of right now there is no word on 1.8-inch packaging but as the new SSDs will only have a z-height of about 7mm the packaging options are pretty wide open.
So where does all this SSD goodness leave Seagate’s platters? I asked Rich that and he stated that the SSD was not a replacement for the traditional disk based media. It was a complement. It enters as an extra level of performance offering; an offering that will be the top of the food chain but still a part of that chain. In reality the SSD hits the field as a “best-in-class” product for the mission critical and performance needy enterprise application while traditional discs will always be the best for bulk storage that does not require the same level of performance and speed.
Rich also told me that we can expect to see Seagate’s enterprise class SSDs before the end of this year. I am eagerly awaiting them to see if the wait is worth it. If these new SSDs are like other Seagate products I am sure they will be.