2009/07/21
TheRegister: Switch to SSD 'for free'? Sandforce explains how
by Chris Mellor
Fabless flash controller developer SandForce has let a pricing hint slip. It will reveal more of its plans and situation at the August 11-13 Flash Summit in Santa Clara. It may also be about to reveal its first supply deals.
SandForce emerged from stealth in April with NAND flash solid state disk (SSD) controller technology that used 2-bit multi-level cell flash and delivered pretty symmetric and fast read and write I/O: 30,000 IOPS and 250MB/sec with either reading or writing of 4KB data blocks.
It dealt with the limited write endurance of the flash chips by minimising the amount of written data with its DuraClass technology. The pitch was based on making cheap but limited life-cycle MLC flash practical for enterprise use, and so undercut single level cell (SLC) flash, which is faster and has a longer life than MLC flash.
Back in April (http://www.theregister.co.uk/2009/04/13/sandforce_launches/page2.html) it had evaluation technology and two products, but now pricing hints have emerged. The company is a bronze sponsor and exhibitor at the August Flash Summit in Santa Clara, and is busy hiring staff.
Productwise, SandForce has the SF-1200 mobile processor - processor being its term for controller - and the SF-1500 enterprise processor. These are both application-specific integrated circuits (ASICs) with firmware.
The SF-1500 does 250MB/sec sequential read and sequential write using 128KB blocks of data. It does 30,000 random read and write IOPS with 4KB blocks. Sixteen flash devices, from virtually any flash supplier, can be used, making it possible for OEMs or system integrators to provide various capacity SandForce-controlled SSDs. SLC chips could be used, but that would negate the making-MLC-enterprise-class pitch.
The SF-1200 has pretty much the same bandwidth (250MB/sec read versus 200MB/sec write) but lower IOPS, doing 5,000 IOPS for random 4KB reading and writing.
Pricing hints have come from an article in Greentech Media (http://www.greentechmedia.com/articles/read/a-chip-to-slash-the-power-in-data-centers/), in which SandForce CEO Alex Navqui said SandForce SSD storage could be free if its total cost, acquisition and electricity supply costs over five years are compared with the similar costs of replaced hard drives over the same period. The amount of saved power cost would pay for the SandForce SSDs.
Navqui uses as a comparison a set of 240 73GB fast hard drives, which have a total storage cost/GB of $3.16 or $50,000 over five years. He says you only need nine SandForce SSDs to deliver the same IOPS and they have a 5-year energy cost of $250. Subtracting their cost from the HDD energy cost gives us $49,750, and dividing that by the number of SSDs (nine) gives us an SSD unit cost of $5,527. These will be the SF-1500 processors, with 2-bit flash chips added. We don't know the unit capacity of the SandForce-controlled SSDs.
SandForce processors are intended to be used by SSD suppliers who do not have or do not want their own in-house controller technology. Given that Seagate and SandForce share a board member, C S Park, and that Seagate has reaffirmed its intention to ship its own SSD this year, there is a fairly good likelihood that it could include a SandForce controller.
IBM provided a supporting quote when SandForce came out of stealth in April but it is questionable if IBM wants to become an SSD supplier. It is perhaps more likely that a flash chip supplier, such as Samsung, might use the SandForce technology and so be able to supply system suppliers such as IBM.
Looking at the Flash Summit programme (http://www.flashmemorysummit.com/English/Conference/Seminar_Session_Descriptions.html), we notice that Xiotech's VP for Technology, Rob Peglar, is appearing in several forums. Xiotech announced SSD support in its Magnitude 3D storage arrays in 2006. It is possible that Peglar's appearance at the event may presage some kind of Xiotech SSD announcement.
The usual suspects - Intel, Micron, Numonyx, SanDisk, Samsung, STEC and others - will also be out and about at the Flash Summit, and we might well expect a mini-blizzard of flash announcements at the show.
It's noticeable that enterprise flash disk developer Pliant Technology (http://www.theregister.co.uk/2008/02/21/pliant_startup_decloaking/) does not have a presence at the Flash Summit. There were rumours of a summer announcement by the company, but nothing has come to pass and Pliant looks to be staying in stealth mode for a while longer. ®
InfoWorld: Adobe offers up more Flash technology to open source
by Paul Krill
Adobe Systems will offer more of its Flash rich media application platform up to open source Tuesday, a move viewed by analysts as reactive to the fierce competition Adobe faces in the rich Internet application space from the Microsoft Silverlight platform.
Also in the open source realm Tuesday, Canonical, commercial sponsor of the Ubuntu Linux distribution, will offer code for the Launchpad software development and collaboration platform in an open source format. The Adobe and Canonical contributions follow by one day Microsoft's contribution of 20,000 lines of device driver code for Linux.
Adobe will make available via open source the company's OSMF (Open Source Media Framework) and Text Layout Framework. Formerly part of the "Strobe" project, OSMF allows for software-based media players to be built based on the Flash platform. Individuals could, for example, add new functionality around the Flash Player.
Text Layout Framework allows users to "to do all the things you want to do with text to make it really cool" on the Flash platform, said McAllister. Sophisticated typography capabilities can be added to Web applications.
The two offerings follow previous Adobe efforts to open source parts of the platform. Previous Flash technologies released via open source have included Flex and its compilers, and the Tamarin virtual machine. Specifications also have been released for streaming formats
"People quite often think that the Flash platform is a closed platform, Adobe-only," McAllister said. "What we're doing is continuing this commitment to making the unique features of the Flash platform open."
Although Adobe insisted its latest open source efforts were not done as any sort of response to Microsoft's Silverlight, analysts nonetheless saw a Microsoft angle.
"It's yet another example of the serve and volley going ( on ) in the RIA space," said Jeffrey Hammond, principal analyst for application development at Forrester. "Adobe and Microsoft are pushing each other hard, and as a result, the state of the art for RIAs is advancing at an amazing rate."
"Adobe is in a race with Microsoft for RIAs, and open source is a powerful way for Adobe to level the playing field considering Microsoft's huge mindshare and adoption among developers," said Melissa Webster, program vice president for content and digital media technologies at IDC.
The core Flash Player and Flex Builder IDE remain unavailable to open source. " McAllister said. "There's code inside the Flash Player that we don't own," such as codec technology, he said. Flex Builder, meanwhile, is built atop the open source Eclipse IDE.
Still, developers have more open source options with Adobe than with Silverlight, said Webster. She also cited Adobe's Open Screen Project as an example of openness.
"Yes, the Flash Player remains Adobe-proprietary, however with the Open Screen Project, developers can write their own servers to stream media to the Flash Player," Webster said.
Adobe's OSMF and Text Layout Framework better enable companies to take advantage of capabilities of Flash 10 without having to understand all the "nuts and bolts of low-level ActionScript calls and functions," said Hammond. ActionScript is the programming language for the Flash platform.
Adobe is working with Akamai to coordinate OSMF with Akamai's Open Video Player initiative. The companies will provide a framework enabling partners such as developers and content owners to build new services with high-end features.
Canonical's Launchpad, meanwhile, lets developers host and share code for free using the Bazaar version control system. Developers now can contribute directly to Launchpad themselves.
"Launchpad is designed to accelerate collaboration between open source projects," said Canonical founder Mark Shuttleworth , in a statement released by the company.
"Making Launchpad itself open source fulfills a long term intention to give the users of Launchpad the ability to improve the service they use every day," Shuttleworth said.
While open source projects are hosted for free on Launchpad, closed source projects can use the service for a fee.
CNet: Moore's Law vs. the cloud
by Gordon Haff
We've been hearing a lot about thinner client devices of late. Netbooks are a hot topic, whether or not they're really a distinct category of device. I've wondered if there might not be a role for a sort of ebook-on-steroids. And Google's Chrome OS, pitched for a browser-centric world, had the digerati all in a flutter a few weeks back.
A lot of this activity reflects a general move away from software that is locally installed and run on a traditional PC to software and services housed on servers out on the network--in the cloud, to use the lingo du jour. It's enabled in no small part by increasingly pervasive networks including wireless ones of various kinds.
However, although cloud computing tracks improvements in networks, it doesn't necessarily sync up so cleanly with the parallel improvements going on in computers themselves. As a commenter put it in a recent post of mine: "The thing that I don't understand about the move to "cloud-based services" is that it seems at odds with Moore's Law. Specifically, devices are going to have more & more processing power, disk space & memory - why would you want to offload processing to the cloud?"
This is a deceptively deep comment and one that touches a lot of basic architectural questions about how we will run software and where we will run it.
One thought is that we're not really running counter to Moore's Law. Rather, we're moving the increased number of transistors that Moore's Law gives us from the client to the server. We're making clients thinner (and therefore more portable, cooler, and so forth) and the servers fatter.
There's some truth in that with mobile phones perhaps offering the clearest illustration.
But, for more notebook-like clients there's a lot of processor and graphics horsepower on the local computer that's going to waste much of the time. And, in any case, telecommunications infrastructure places hard limits on bandwidth for a given time of place, but we can dial up and down our local compute horsepower by selecting devices with different characteristics. So it makes more sense to favor local processing much of the time.
In fact, the fundamental thing that thinner clients and cloud computing tackle isn't really the movement of computing off the client but rather the movement of "state" off the client--which is to say data, applications, and customizations specific to a given user.
As a practical matter, most clients still store some amount of state. In the days of old, terminals didn't store anything locally. Sun's Sun Ray line comes closest to replicating this experience in modern thin clients. However, even browsers store cookies and can be configured with extensions and plug-ins that will vary from one installation to the next.
And, for most purposes, this is probably a reasonable enough state of affairs. Our personal devices are personal anyway; we just want to get away from having to load and manage custom software for each individual task that we want to do. Shared, public clients are a different matter, of course. However, in this case, a lowest-common-denominator software load (such as a browser) is typically sufficient.
There is clearly a lot of work left to do and battles, both technical and political, left to fight to arrive at the best architectural models and programming practices for this new generation of client-server computing. For example, do "rich Internet applications" live in the browser a la Microsoft's Silverlight or is a separate framework such as Adobe's AIR a better approach? Where do .NET and Java fit in?
These (and many others) are not small questions. Application writers need to understand at a very granular level the environment for which they're writing. And there is very much a tension between richness of the client experience and the degree to which we can standardize and simplify that client.
