Atrato looks for cash to fill out sealed canister storage pitch

Xiotech may have talen the public perception lead in shipping drive arrays based on self-healing and sealed canisters of disk drives, but Atrato is shipping the same Seagate-derived technology into the video surveillance market.

It launched its product, the Velocity 1000, last year, but its public profile is near stealth-mode outside its immediate set of resellers and customers.

The V1000 is based on sealed canisters of twenty or so 2.5-inch hard drives, with up to 160 drives in a system. Atrato software and technology with spare drives means that the arrays are essentially self-healing - a 3-year, maintenance-free warranty is offered. The use of multiple 2.5-inch drive spindles accelerates I/O. Atrato targets the V1000 at video collecting and video delivery businesses, such as video surveillance and cable TV head-end functions which need the multiple simultaneous delivery of hundreds of video streams, also other data-intensive applications needing high drive array performance from a compact unit.

It says its V1000 can do 16,000 random read IOPS at 4k block size, has a bandwidth of 1.4GB/sec when doing random reads at 512k block size, and can output 3,700 content streams at 3.75 Mbit/s. All it needs to do this is 3U rack unit.

Xiotech's sealed drive canister technology comes from Seagate, from work done by Steve Sicoloa's advanced architecture group. The overview history goes like this: a storage start-up called Xiotech was founded in 1995. Four years later it was bought for $360m. Both Dan McCormick and Jonathan Hall were involved in Xiotech at that time. Seagate later acquired a storage technology development group, led by Steve Sicola, from Compaq.

This group became Seagate's advanced storage architecture operation and developed the self-healing and sealed canisters of drives concept as a way of reducing the consequences of drive failures and cutting maintenance costs. Spare drives and clever software in the sealed canister meant that it could work around failed drives, and operate as if it was one big disk drive.

Eventually, in 2002, Xiotech was floated off by Seagate. McCormick left Xiotech in 2003 and he and Hall started Sherwood Information Services, which renamed itself Atrato in 2004. They used a sealed canister idea to develop their SAID architecture, meaning a self-healing array of independent drives, and the V1000 product was launched in March, 2008.

Hall said: "The self-healing and self-maintaining architecture in Atrato’s velocity subsystem was developed completely within Atrato and is unique." Hall's background was: "Enterprise disk drive design with an in depth understanding of rotational vibration directly applicable to Atrato’s patents and SAID enclosure design."

Seagate sold off its Advanced Architecture group to Xiotech for $40m in November, 2007, when it found no takers from its OEMs for products using the sealed drive canister concept. Xiotech then developed its Emprise line using the technology and launched it a month or two after the V1000 hit the streets.

Now Xiotech is apparently doing well with Emprise, while Atrato appears to be motoring along but not making any waves. It has raised $18m in venture capital funding and the line-up of investors and advisors include:

- Jesse Aweida, Aweida Venture Partners, founder and former president/CEO of StorageTek
- Tom Porter, formerly CTO, Seagate and an IBM storage executive
- Gary Gentry, SVP Maxtor, and then Seagate
- Dick Blaschke, an IBM and EMC veteran.

There is a lot of Seagate influence here.

Atrato thinks it will earn more than $10m in revenues this year and is looking to raise another $15m in a D funding round. This is not the best environment to be chasing VC funding. However, Atrato's product pitch of lower maintenance costs for a high-performance array should strike a chord with cash-strapped customers.

The company hasn't publicly said what it will do with the money but engineering development around 2.5-inch format solid state drives might be involved as it is talking about tiered arrays. The original SAID concept involved identical drives so any tiering will involve development costs. There is also the introduction of 6Gbit/s SAS to cope with.

Money could also be needed for channel development with alliances being pursued to integrate Atrato's box with application and other hardware elements needed to make up a complete customer system.

Although Atrato is named after the fastest-flowing river in the world, located in Colombia, it is nowhere near producing a river of cash. Mainstream arrays using 2.5-inch format disk drives have not yet appeared and the combination of two newish technologies, the 2.5-inch drives and the sealed canisters, means that customers need educating and convincing about the benefits of the product. It just has to get more sales and build momentum to take it forward. Good luck. ®