Short note on a story John linked to on InsideHPC.com
Should HPC vendors go private is the question. Of the three vendors listed, two of them are HPC vendors, the third is a general vendor with a few HPC sales.
Ok, who are the HPC vendors? This is a good question. I won’t give an exhaustive list, but the usual suspects are on that, where they derive all or most of their revenue from HPC. Cray, SGI, Panasas, SiCortex, Voltaire, … etc. Then you have the vendors with HPC sales/development groups (large or small) where HPC doesn’t make up the most or even a majority of the revenue. This would include IBM, HP, Dell, Qlogic, Cisco, and I would include Sun in this list. Then you have the vendors who really have little HPC footprint at all. This is most of the storage vendors, and despite their efforts, companies like Microsoft. This group are companies for which HPC may be a (permanent) loss leader, and is being executed as a tactic in a larger strategy, for competitive reasons, or for reasons relating to “check boxes” which allows the company to claim that they have HPC capabilities, in order not to be excluded from various business opportunities.
I haven’t included smaller shops like mine or any of our many partners such as Tsunamic Technologies in these lists, but we could.
Most in the middle and last group are public companies, and should remain so IMO. They have many lines of business, and a critical focus or decision making based upon one line of business impacting all lines only makes sense in the context of that being the majority of business. That is, their tails shouldn’t wag those dogs. In fact, it might become meaningful for some of these companies to, at some point, jettison or shutter their HPC portion of business if it no longer aligns with the corporate direction. I am not implying this will happen. It has in the past with other companies. SGI famously got out of clusters in January/February 2001. I left them in March 2001 (as a direct result) as I thought clusters were the future of HPC (but I was too chicken at the time to form Scalable Informatics, which I did 16 months later).
Should SGI go private? I don’t know. Not sure if this is a good idea. This would reduce their likelihood of being able to raise additional capital going forward (IMO), as the risk would be concentrated in fewer hands, and they would be (likely) more demanding on the management to perform and cut lossy business lines. SGI would lose some of its capability to retain excellent staff by offering stock options. If the employees can’t trade it, why would they want it … it would have no value to them, and therefore wouldn’t be usable as an incentive.
Should Cray go private? I don’t think so, unless someone outright buys them. Same issues as above, though Cray has a different sales model, in large part dictated by various government agencies around the world. Its a longer sales cycle, and it can result in good financial performance, especially if they start exploring regions of the market that haven’t been well served by existing technology. This is something Cray can do, as they do build their own stuff.
Should Sun go private? No. This one is obvious. They have too many lines of business. I might think that they would want to divide the company between hardware and software divisions, and have each follow its own path.
Going private isn’t a panacea, or a solution to anything. It simply is a way to hide real state from the public eyes. Chrysler LLC doesn’t have to report its financial state to any external non-investor group. So no one knows precisely how it is doing outside of a very limited group. But has this made their lives easier? I don’t think so. I don’t think it has made them harder either.
In HPC many of the private companies are looking to go public. OTOH, if the stock is cheap, it is well worth the public companies to buy it up. As much as possible without going private. This will reduce the supply of equity in the company available to the public, and in doing so, hopefully allow them to have a higher per-share price. All they have to do is perform.
That is the issue. If they are going to perform better now or in the future, buying stock, or buying back their debt is a good thing. If they are not going to perform better now or in the future, they should hunker down and wait for the market to rebound as a whole. A rising tide lifts most ships. Not the ones securely sunk or sinking, but most ships.