M&A: Microsoft buys the *assets* of Interactive Supercomputing

As seen on InsideHPC, John West notes that the assets of Star-P were purchased by Microsoft today.
Parsing of words is important. The phrase “acquired the assets of X” means that the IP was purchased. John points to the blog post where Kyril Faenov mentions that some of the staff will work at the Microsoft Cambridge site.
This is sadly, not a great exit for Star-P.
Acquiring assets usually means the choice has been to shut down the company, and auction the bits off, or find a buyer for the distressed assets and then wind down the rest of the organization that doesn’t go with the assets.

I am not sure precisely what happened here, but its not looking that great of an exit for a startup.
Specifically, Star-P was effectively a cross platform tool. Our partners, Tsunamic Technologies have a Star P on demand service, running on a Linxu cluster. I think it is highly unlikely that Microsoft will continue the development of the Linux toolbase. The acquisition of assets gives them an easy way to avoid having to worry about that conversation.
In fact, this article does confirm this direction:

Microsoft says it will not continue developing Star-P (ISC???s flagship product) beyond version 2.8 which was released earlier this year, and that version 2.9 that was released to a few customers in Beta will not be released for production use by customers. Active Star-P customers who are using earlier versions of Star P were granted the right to upgrade to 2.8 by ISC prior to the close of the transaction.

But thats not all …

Microsoft did not acquire the customer contracts between ISC and their customers but says it will provide technical support to active customers through the longer of their existing support contracts or 12/31/2010.

This suggests my thesis of them splitting the company assets and liabilities, shuttering one aspect, and absorbing another. The Partners page confirms this:

3. How does this acquisition impact my current ISC Partnership (contract, alliance management, etc.)?
Since Microsoft has not acquired ISC???s partner contracts – but has acquired technology assets – your existing ISC alliance management relationships and contracts remain with ISC.

But … the Interactive Supercomputing website now redirects to a Microsoft site, with a discussion of the “transition”.
More to the point

2. How will support for ISC customers be impacted
* Customers will get a new support phone number, email and URL to access support but support should not materially change.
3. Who does the ISC customer need to contact for customer service and support?
* Notice of the new phone, email and URL has been sent to customers.

John was right about the deadpool, but possibly for a different reason than I think he intended. Yes they are gone, but a technology available in market, which was/is valuable cross platforms, is being removed from the market. Sure, you can argue that its going into future Microsoft products, but you have to ask yourself … how is this a win for high performance computing users, the (vast) majority of which are not windows HPC users?
John opined that Rapidmind should also join the deadpool. I am not so sure of this. You can still buy Rapidmind, their website is not gone, and their customer contracts will be honored. Nothing was removed from the market.
Cilk Arts is also not a deadpool candidate. For the same reasons as Rapidmind isn’t IMO. You can still buy it, it is still available, and the team is still working on it.
One could easily argue (and I think I did in the past) the the Lustre/CFS acquisition, being an asset acquisition, was not good news for CFS. They were (to a degree) in the deadpool … as much as an open source tool, and company around it, can be in the deadpool.
To Rosario C (former colleage from SGI days), David, and the rest of the crew at Interactive Supercomputing … I hope this works out well, I am sad to see the product removed from the market.
There is no doubt that this is a tough economy, and companies with good products are running out of cash while investors hedge bets, or hold back on subsequent rounds. I’d bet a reasonable beverage at SC09, that this is approximately what happened … sort of a software version of a SiCortex, though Microsoft acted before the company shuttered to guarantee others didn’t get the technology at auction.
Note: and this is important … I am not impuning Microsoft’s motives … they saw a technology and grabbed it. I am pointing out that it would be hard to classify this as a win for the investors, employees, and customers of Interactive. Probably a loss for the investors, and a draw for most of the employees. A definite loss for the customers. Especially if they built a business or process dependency around a closed source tool, that suddenly vanished, without a reasonable alternative.
Anyone wanna bet that Microsoft is looking at Mathworks closely? Though Mathworks would command a considerable sum.
Yeah, I’d call it (Interactive Supercomputing) a deadpool candidate. I’m not happy about another HPC company shutting its doors.

6 thoughts on “M&A: Microsoft buys the *assets* of Interactive Supercomputing”

  1. Another reason I’d see Rapidmind and Cilk continuing to exist is that Intel’s business drivers are different than MS. It’s in MS’ best interests to snap up and offer differentiating technology on their own OS platform and brand. Intel as a hardware vendor has a “Come one, come all” perspective, anything that sells chips and helps people leverage multicore is good for them.
    I wouldn’t be surprised to see MS buy Frontline and incorporate it into Microsoft Solver Foundation. They’re already pretty tight.

  2. I’m still trying to figure out what “deadpool” means. I think it should mean that the business has ceased to exist under its original identity. This is the most basic definition, but has the benefit of not having to parse the technology situation and make judgment calls on the value of what a company does with the technology it acquires after the acquisition.
    By the way, I appreciate you parsing this announcement so closely, though I’m still optimistic that this is a reasonable thing for the HPC community. I’m not happy either about less diversity in the ecosystem, but I am hopeful that the technology will survive in a way that it did not with PeakStream.

  3. To me it looks like Star-P is dead, it’s been assimilated by the Borg and as such is useless (unless MS start supporting Linux, heh).
    Amused to see “Viewed 6666 times” on this article, an order of magnitude worse than the beast ? 😉

  4. @Chris
    Sad to see it go, it was absorbed like Google absorbing PeakStream.
    On the article counters … I seem to have a fairly good/active readership from what the statistics tools are telling me. Spiders/crawlers aren’t even close to a majority, most are rss feeds and other things.
    A rather large readership from Wall Street and financial types in the US, UK, and other locations (hi folks! glad to see you stop by! feel free to comment!), and multiple other significant high performance computing and storage vertical market users.
    This is good of course, and I welcome more readers, and those with a comment or thought to share.

  5. Some questions come to mind.
    1. Why Mathworks did not get them?
    If “Interactive SuperComputing” were in distress and looking for a buyer, why did not Mathworks get them?
    2. If they were successful and expanding, then the price they commanded may have been too high for Mathworks. But not for Microsoft. Microsoft does not buy junk technology. No company does. They are not doing it for charity either. I wonder if success of Start-P has simply gone too far for them to witness. And they jumped right in to (1) curb the spread and (2) assimilate it themselves. Although it is so extremely vague and far-fetched on how they can actually do it, I am inclined to think it is purely the former.

  6. @Nick
    I am guessing that they were working on a deal with Mathworks at one point. Many smaller startups have acquisition as their exit strategy these days, and the approach is to build upon a successful product offering with something (hopefully) perceived as valuable.
    The problem I saw with Interactive is, while their technology is cool, its wasn’t a “killer” technology that everyone who has Matlab needs. The value proposition existed, but for a limited set of their customers. So my guess is that Mathworks, if it made an offer, made a small one. And Interactive likely burned through its cash, and had to choose between shutting down, or selling assets.
    This is a guess of course. If they were successful, they wouldn’t have had to sell. I have a sense it was rushed, as in one deal fell through. Pure speculation on my part.
    The acquisition fits in with Microsoft’s expansive strategy. They identified Matlab as an important technology, and my guess is that they will try to take ownership of the ecosystem around it.
    As for Microsoft not acquiring junk … Microsoft isn’t known for providing good ROI to original investors for companies they buy. They drive hard bargins. It helps them that they have a number of monopolies they can exploit. I don’t know that their investments have done very good by them.
    Overall, I am concerned about Microsoft’s commitment to the space. Buying Interactive’s assets may get them some capability. Buying Mathworks would get them much more, albeit at a much higher cost. One I am fairly sure Microsoft wouldn’t be willing to pay.
    Interactive was cheap (relatively speaking), and buying distressed assets is often a good risk if you can deploy those assets or leverage them somewhere else. I think Microsoft is going after the envelope strategy. But unless they are committed to the space, this could be a problem.
    I know of one organization that had a business product offering that depended upon Interactive’s Star-P product. This is the issue with closed source codes, once the vendor goes away, your risks increase massively. This is true regardless the size of the organization (which is why Sun’s revenues are cratering right now). How can you replace that line of business if the underlying technology product goes away due to the company disappearing?

Comments are closed.