A good question

John at the always interesting InsideHPC asks a very important question, that, oddly, I think I can answer.

The overall article is on the SGI salvage of LNXI assets. John’s question was

I’m sincerely not being snarky here, but how, exactly, is SGI showing signs of improvement?

They are not. You can see it quite clearly on the financial charts

The company’s stats can be read from the Yahoo page, go ahead and click the picture and you can see it.

Their market cap is less than $200M. Price to sales is under 1. This means that their sales are being effectively discounted. Inbound revenue, is something like $440M. They have $85M in debt, $33.2 M in cash. Net cash flow is outward, to the tune of $49.2M over the last twelve months.

You can compare these to companies doing well. Even if you scale by the market cap, the other companies have significantly different cash flow and value profiles.

Moreover, if you look at what SGI did relative to LNXI, they basically purchased LNXI’s assets, which had secured its debt.

These assets included LXNI’s system management software products, intellectual property and intellectual property rights and were acquired by the Collateral Agent through a non-judicial foreclosure on behalf of the Secured Parties in partial satisfaction of indebtedness owed by LXNI to the Secured Parties. As consideration for the sale of such assets, the Company issued 390,000 shares of the Company’s common stock to the Secured Parties pursuant to the terms of the Stock Purchase Agreement (as defined below).

LNXI borrowed money. I explained why previously. The problem was that LNXI was not likely able to service this debt, due in large part to cash flow instability, caused by non-payment or extraordinarily late payment on the part of their large government customers.

As I am fond of reminding people, there are consequences to every decision you make. Some of these decisions and actions have some un-anticipated consequences. Driving the vendor out of business is not likely the goal of every government procurement. I wonder what the purchasers of all those nice big LNXI systems are thinking now.

Onerous T&C are great to force vendors to deliver what they promise. It will also effectively force all the innovative vendors, the ones that take risks, out. Well, they can take the risks, and do well if they succeed, but there is risk. The T&C we have seen suggest that not only do the purchasers want a low price, they want no risk. Drive enough small companies out of the market, and you are going to get a uniformly bland bit of me-too clusters. Which is where things are largely headed.

But I am digressing.

SGI bought assets at a fire sale, by selling the purchasers some SGI stock.

Is SGI doing well? To answer John’s question, as I alluded to above, they are not. Assuming around 1500 employees (rough guess), their revenue per employee is about $293k. It should be closer to $400k for a company doing well. I am not concerned about their debt load, they should be able to service that.

SGI is in direct competition with Dell. And HP. And IBM. And (to a degree) Sun. And the second/third tier. My day job is not in direct competition with them most of the time, though we do see them every now and then. For those who don’t know, I worked for them from 1995-2001, leaving after getting disgusted over their exiting the Linux cluster market … for the first time … in 2001.

Trying to out-Dell Dell is not what I would consider a wise business move. Nor HP for that matter. Both can move huge volumes of systems, leverage economies of scale that SGI simply cannot. Moreover, as you can see if you go to Dell’s site and start configuring things, Dell can drive the prices of units below their BOM costs for a while and survive. This has the net effect of removing oxygen (and cash flow) from competitive sales. Dell can afford this, as they sell millions of units. So can HP. And IBM.

SGI needs differentiation. Its Altix systems (well the original ones) were differentiated. But they are also quite expensive. And based upon Itanium’s.

Do LNXI’s assets differentiate SGI? I don’t think so. There are lots of products/projects out there that provide command/control over clusters.

This is not to say that SGI cannot differentiate. They have had (and still have) some very bright people. The issue is that the HPC market of old has gone away, and the new market requires very different business models, product mixes, and so forth.

They exchanged something like 300k shares, at $18/share for the IP. About a $5M transaction.

I think they even scored an investment at the same time. So from SGI’s point of view, this asset acquisition was effectively a no brainer. Give them stock (no cash), get IP. And get to hire some really good folks (LNXIers were no slouches).

Unfortunately, in this market, there are no silver bullets. There are no killer apps. There is nothing you can do to easily make people start buying from you versus the competitors.


Because these are commodity machines with commodity processors, and minimal differentiation.

Well, most of the machines in this market are. Some are differentiated by what they do, how fast they do it, and so on. Still, it is hard to convince people to buy better products than what they are buying now. We have struggled with this in the past with JackRabbit. Though the rate of orders has increased dramatically, so maybe our message is being heard. Or word of mouth.

The point is that a box from SGI is about the same as a box from Dell as a box from HP as …

So why get it from SGI?

For LNXI and other small HPC vendors (like the day job, but unlike the rack-n-stack folks), it is easy … our systems are more highly tuned and better fits to customer problems. For the large vendors, it is easy, they can deliver huge volumes of consistent machines. Not likely at as good a price point as the smaller vendors (with smaller cost structures), unless they get really hungry.

SGI leadership’s job is to convince customers that they have something that the others don’t. Don’t convince Addison Snell (an SGIer of about the same time period I was) at IDC. Convince the people who buy things. There is marketing, and there are results. We have seen the corporate marketing. We need to see the results to be convinced that SGI is “really” back. I don’t think they are. Moreover with their cash outflow rates, they may have issues in the future.

As someone who runs a small business, I can tell you that watching/controlling cash flow is critical. Part of this is keeping spending in check, part of it is winning business. You can only do so much of the former, you need to do the latter to grow the business.

Growing a business is hard. We have been doing it for half a decade. You have to learn, adapt, try, fail, succeed. SGI can’t do more of the same stuff it has done in the past. It has to grow/adapt.

You can choose to sink, or swim. Then you have to take action to make it happen. Talking about swimming isn’t swimming. Its talking. The only thing that matters is the swimming.

Viewed 7678 times by 1735 viewers


2 thoughts on “A good question

  1. Joe – great post. I enjoyed the financial lesson, and I think you are spot on in pointing out the holes in the failing business model of old line HPC.

    Sadly, I also think you are right about the direction that onerous T&Cs are pushing the industry. For that one, though, I find it hard to imagine a fix. Big government purchasers have made such a big deal out of getting maximum performance for minimum price, and have been so successful in framing that discussion in only financial terms, that giving the industry enough profit to spur innovation would amount to admitting that the government is starting a program of paying more and getting less. Without strong leadership from the top, no mid-level government manager is going to be able to pull that one off.

Comments are closed.