Fusion-io makes no bones about it. It wants to get in the data warehouse space. But to play in that space, it first had to deliver a solution that scaled to over 10 TBs. Today Fusion-io’s CMO Rick White and I continue our interview series as we discuss why delivering on the 10 TB benchmark was so critical to both Fusion-io and its prospective data warehouse clients and why it is going directly to the channel to deliver these solutions.
Jerome: So tell me about the Fusion-io Octal.
Rick: The whole concept around the ioDrive Octal is really very different than what we are doing now. We have had a lot of success in scale-out but not in scale up.
ioDrive Octal Flat Shot – source: Fusion-ioCustomers that are running a SAP HANA appliance are a great example of those who are running a data warehouse that would love to run it all in memory.
We announced the 2.5 TB ioDrive Octal and we were excited about it. But it was a lousy product to announce if you think about our customers which are scale-out customers as they could not use 2.5 TBs.
The 5 TB ioDrive Octal that we soft launched later was even worse. It was too big for scale out customers and too small for scale up customers.
The sweet spot for these scale up customers – data warehouse and heavy data analytics – is 10-20 TBs. 10 TBs gets you a seat at the table and, at 20 TBs, you can support most everything as all of their data sets tend to fit within 10-20 TBs.
So with a single Octal you are at 10 TB and with two Octals you are at 20 TBs. There is a 1U HP server that gives us two 10 TB Octals in a 1U form factor. So now you can take these Octals and put them in any HP, IBM, Dell or Supermicro server and bundle it with our data migration software and you literally have a turnkey flash storage appliance.
Jerome: So what is your go-to market with this?
Rick: It is simple: we are going to the channel. 75% of revenues for HP and IBM – it is actually closer to 76% – come from their channel. So we are engaging their resellers and taking it upon ourselves to train them.
After you have been in the industry long enough you know that it is not just, “Hey look! We have signed up an OEM and we are going to make tons of money. Wahoo!”
If you look at HP and IBM, we sell a lot of flash through them even though they already sell Intel, LSI, Samsung, and the STEC MACH line. In spite of that they still sell our product.
In our case, we recognize that we have to have a direct sales force to create the pull. We need a team that goes out and trains, engages and does joint marketing with the channel.
We are going to give the channel solutions. These integrators do not have large software teams that are dedicated to development or R&D. Using us gives them a turnkey software solution that they can load and integrate inside a server.
In this way they can buy a HP server and an ioDrive from Fusion-io and we can then help train them so they have a turnkey appliance to go out and sell as a VDI appliance or configuret as a database accelerator or whatever.
In Part I of this series, Rick discussed how server-based flash is poised to change the enterprise.
In Part II
of this interview series with Fusion-io’s CMO Rick White, we will
discuss why this decoupling of I/O performance from storage is necessary
and why this creates a new tier of memory as opposed to a new tier of
storage.
In Part IV of this interview series, Rick and I discuss why Fusion-io is opening up its virtual storage library (VSL) APIs to developers.
In the final Part V of our interview series, Rick provides Fusion-io’s takes EMC’s Project Lightning (now known as VFCache) and the gap that persists between SSD providers and Fusion-io’s ioMemory.