Posted: 01/25/2012

(repost:  http://www.fusionio.com/blog/a-few-facts-on-fusion-io-customers/)

At Fusion-io we view every customer as a partnership—we are successful when our customers are successful. Consequently, we continually gather information and put it to work to help us better serve our rapidly growing customer base. We thought we’d take some time to share some interesting results from some of our research:

• Roughly 33% of Fusion-io customers are willing to be a resource in educating future customers in Fusion’s products and services, so don’t hesitate to ask us for a reference!
• Over two-thirds of our customers are database users, including Microsoft SQL, SAP, Oracle, MySQL, Postgres, and more.
• A growing number of customers are utilizing Fusion-io in their virtual environments, including both server virtualization and VDI.
• 70% of our customers buy at least two ioDrives with each purchase, with close to 20% buying four at a time.

Interesting in learning more about Fusion-io enterprise storage? You can read our case studies, check out the latest news, or jump straight to our product page.

• bugs are ironed out sooner,
• the products are more reliable,
• the products become more widely known
• the products are supported on more platforms
• the products are incrementally cheaper to make

It is to be hoped that at some point in the future – as the innovation curve flattens – and the technology creator’s brand strengthens and the product becomes a sticky standard supported by compatible 3rd party partners – the margins in the product itself and in the channel mix may change for the better. (Licensing dealstoo are another possibility for extracting more profit from high volume oem customers.) There are no guarantees in any competive market but that’s my way of trying to make long term sense of what’s going on in some hot spots in the SSD market today.

(full article here:  http://www.storagesearch.com/ssd.html)

Kevin,

Your recent comments regarding FIO are off the mark, in my opinion.  The competition in the enterprise is, so far, non-existent; ask anyone at the end-users or even at the mftrs.  Can you find EVEN ONE data center to stand up and say we have a 2nd source w/ Virident?  OCZ?  TMS?  Or you can go direct to LSI, OCZ, Virident:  where has anyone developed even one relationship (OEM or end-user) or sold even one device to a major data center?

You keep saying there is pressure on FIO — but where is it coming from?  I think it comes from Violin, Exadata, etc — perhaps Solidfire, Nimbus.  But even Kaminario and Schooner use FIO inside  — so the appliance v. PCIe argument is less meaningful.

I think FIO shot itself in the foot this qtr — as a young company it really does not know how to manage expectations yet.  More importantly, the market is blue ocean and highly unpredictable.  But in the meantime, they are growing (incontrovertibly) to $100m per qtr.

I think you will agree that they are looking at something north of $350m for the FY.

And, at the same time, they are sucking the oxygen out of everyone else’s fire.  ioTurbine and ACM are truly disruptive — it takes time to get the message out, show the value, funnel the sales efforts.

The margins will come back — we both know this — and when they do, FIO is $40 and beyond.  They really ARE the next VM, DDUP, Riverbed — you name it.

I am a former employe and long FIO.

As a comparison, OCZ is at 81% and STEC is at 87%.  The reason is simply there are not enough shares of FIO to go around.  That will change next month when an additional release of shares comes online and institutions can purchase without triggering SEC-required reporting

Decide for yourself if Thomas or Scalia’s dissent holds water.

About My Finances

Over the past several days, my personal finances have been distorted into a grotesque caricature by the mainstream media, pundits, and other people who can count.  I am writing to you to set the record straight by explaining my finances in terms the American people can relate to.Dear American People:
Let’s say you bought a bottle of Chateau Lafite Rothschild 1982 for $5,000.  A couple of years later, what do you know, you sell that same bottle for $10,000.  So you just made a profit of $5,000 through your own hard work.  How much of that should you pay to the government?  I’d say fifteen percent.

Now let’s say you have a fellow mowing the lawn at your 7,000 square foot home in La Jolla, and he turns out to be an illegal.  You say, “No way, Jose” (Jose is actually his real name) and send him packing.  He doesn’t deserve his full paycheck, since he lied to you in Spanish, but it wouldn’t be fair to give him nothing, either.  So you pay him fifteen percent.

Now let’s pretend the United States of America is like one big restaurant.  Not a fancy restaurant, mind you, but one that only gets two Michelin stars.  And let’s say that you order a meal of Beluga caviar, white truffles and gold shavings, washing it down with your favorite beverage, Chateau Lafite Rothschild 1982.  The bill arrives and it’s quite a hefty one for a working stiff who only made $375,000 last year in speaking fees.  (That’s right: minimum wage.) So when it comes to toting up the bill, how much should I tip the waiter, who in case you’re having trouble following this metaphor is the IRS?  You got it: fifteen percent.

I think I’ve now shown, using these real-life examples that everyone can relate to, that no one should ever pay more than fifteen percent on their taxes.  If you have been paying more than that, you should get rid of your loser accountant pronto.  That’s another thing I have in common with regular Americans: we like firing people.

So – now that I’ve laid it out in simple terms that even you can understand, do you agree that you and Mitt Romney have a whale of a lot more in common than you thought?  I’ll bet you ten grand you do.

Au revoir,

Mitt

(stolen from Andy Borowtiz)

(REPOST:  http://www.fusionio.com/blog/auto-commit-memory-cutting-latency-by-eliminating-block-i/o/)

Posted: 01/18/2012

Last week’s demonstration of one billion IOPS showcased a new paradigm for storing data through Fusion-io Auto-Commit Memory (ACM). ACM isn’t just about making NAND Flash storage devices go faster, although it does that too. It’s about introducing a much simpler and faster way for an application to guarantee data persistence.

When Simplicity Meets Speed
For decades, the industry norm for persisting data has been the same – an application manipulates data in memory, and when ready to persist the data, packages the data (sometimes called a transaction) for storage.  At this point, the application asks the operating system to route the transaction through the kernel block I/O layer. The kernel block layer was built to work with traditional disks. In order to minimize the effect of slow rotational-disk seek times, application I/O is packaged into blocks with sizes matching hard disk sector sizes and sequenced for delivery to the disk drive. As Linux block maintainer (and Fusion-io chief architect) Jens Axboe points out, most real-world I/O patterns are dominated by small, random I/O requests, but are force-fit into 4k block I/Os sequenced by the block layer to match the characteristics of rotating disks. Note the number of steps in this pathway – each one contributes to latency. Even more steps are introduced in this pathway when the block storage device is at the other end of a network, behind various bus adaptors, and controllers.  As long as memory is volatile, this type of I/O pathway will be the norm.

But, what if an application could designate a certain area within its memory space as persistent, and know that data in this area would maintain its state across system reboots?  This application would no longer have the burden of following the multi-step block I/O pathway to persist that data. It would no longer need smaller transactions to be packaged into 4k blocks for storage. It would just place selected data meant for persistence in this designated memory area, and then continue using it through normal memory access semantics. If the application or system experienced a failure, the restarted application would find its data persisted in non-volatile memory exactly where it was left. To illustrate, how much faster could real-world databases go if the in-memory tail of their transaction logs had guaranteed persistence without waiting on synchronous block I/O? How much faster could real-world key-value stores (typical in NoSQL databases) go if their indices could be updated in non-volatile memory and not block while waiting on kernel I/O? That is the simplicity of Auto Commit Memory. Itreduces latency by eliminating entire steps in the persistence pathway.

Addressing Both Halves of the Problem
Block storage benchmarks such as throughput and IOPS are certainly important, but only address half of the problem. The other half of the problem is the work the application and kernel I/O subsystems must do to package and route data for storage.  Applications can be accelerated by addressing either or both halves of this problem. However, note that, at some point, the overhead incurred by packaging and routing data through the kernel block storage subsystem will become the bottleneck. Breaking through that barrier was the goal of this technology demonstration. Give applications the software mechanisms to avoid this block storage packaging and routing latency and complexity. Let them spend more time processing data in memory, and less time packaging and waiting for that data to arrive at a block storage destination.

Fusion-io does indeed make very fast block I/O devices. What’s most exciting about last week’s demonstration for us is looking beyond fast block I/O devices to show what is possible when you address this this fast device as memory rather than block storage.