InsideHPC on SGI results, and thoughts on industry trends

John West at the excellent blog points out in a short note that

So, the company managed to lose money while getting 25% more orders, and it has significantly reduced its capability to ship product. Now, I???m no financial professional, but those don???t seem like healthy indicators to me.

A number of obvious points about this … the economy has been under pressure. Customers have been buying less. SGI reports that revenue has increased to $93.9M from $79.1M in the third quarter. How much of this is normal seasonal variance (government purchase cycles) versus an actual increase in bookings (e.g. new deals that haven’t been worked on for a while). All you need is one large mega-government deal to bias results one way or another, and make the actual business analysis harder. Say for example, someone buys the next Columbia supercomputer from you and you can recognize revenue in that quarter. Basically the issue of revenue recognition timing (when you can say you actually earned the money) is a slightly fungible game. If the CEO calls up the big customer and asks them to delay payment to next quarter, or pull in payment to this quarter to help the financials … sometimes this works. Specifically, it may be hard to discern real revenue events that occur in the 90 day quarter from payments pulled into this quarter, from large mega-purchases.
One of the better indicators of healthy companies is a decreasing loss making if they are on the negative side (in the red) of the balance sheet. It looks like loss making increased by $16M or so … about a 16% increase. This is year over year. This said, the fourth quarter operating loss was $28.6M versus third quarter of $40.6M. About $12M better over a quarter.
Some portion of this is due to the reduce salaried load after their last layoff. The analysis that follows is a ballpark order of magnitude guess, but I am guessing that the order of magnitude and first digit may be correct.
Take 100 employees making $100k/year. This is $10M in yearly salary. Burden this with a 35% benefits package. So $13.5M/year. Per month, this is about $1.13M in salary/benefits. 3 months per quarter, so reducing 100 people saves $3.4M over this quarter. Likely there was some severance as part of the package. This would come off the expenses portion of the balance sheet, and reduce the effect of the savings.
So of the $12M better than last quarter, it looks like about 1/4 of that is due to a layoff (ballpark, guess …). 3/4 of it could be due to some of these mega-deals popping up.
SGI is basically a cluster vendor these days, with 1600-ish people, some IP, and mostly an external outsourced product portfolio (lots of people can sell the same things as they do, from the same vendors). In this market, they are fighting Dell, HP, IBM, and numerous others. This is a non-trivial task. Dell can (and we have seen this up close and personal) suck all the oxygen out of a room if it really wants a deal. So can the others. Margins on clusters are thin and dropping.
I’ve mentioned this before, but it is worth mentioning this again. If, in the drive to outdo discounts, the various purchasing managers manage to drive vendors out of the market, thus reducing choice and competition … this is good for the purchasing managers (and by extension the people who use and fund these systems) … how? Someone, somewhere needs to start taking a long term view of things. Or stop complaining that vendors leave a market (and therefore their choice is reduced) when the vendor can’t make money at it anymore.
SGI has challenges ahead of it, no doubt. When they left bankruptcy, I opined that the competition against Dell and others would prove difficult. There was blood in the water, and the predators smelled it. It looks like SGI hasn’t regained its footing, and the predators aren’t done yet.