HPE, in a surprise move, has paid $275 million to acquire former Silicon Valley giant SGI. This equates to $7.75 per share, a premium of 30% over what SGI shares were trading at by close on August 10. What are the reasons for this acquisition?
Now, it must be said that this is not exactly the famed SGI of the 1980s—which, due to the advanced graphical capabilities of its IRIX operating system and custom graphics cards, dominated the fledgling Hollywood CGI market. SGI drove the tech behind Jurassic Park, Terminator 2, and Men in Black, to name but a few. In fact, the company was worth billions before the post-millennium dotcom bust. However, a few management disasters, the rise of the PC, and cheaper but equally capable graphics cards from the likes of NVIDIA and ATI all but destroyed its position, leading it to file for bankruptcy in 2002. An excellent set of articles from VizWorld maps out the reasons for this giant’s fall from grace. After bankruptcy, SGI was acquired by Rackable and started its fight back, carving out a fairly successful but niche market in the HPC arena.
It is for this reason that HPE is buying SGI—or, to be more exact, its NUMAlink interconnect. The NUMAlink interconnect is what powers SGI’s high-end NUMA system, the UV 3000. This is designed to run resource hogs like SAP HANA without breaking a sweat, with the resilience an x86 architect could only dream about. (It will be interesting to see what happens to the SGI Dell tie-up to sell this product.)
These machines are serious beasts, but HPE already has a pretty decent HPC product with the Superdome and a decent interconnect with the technology it acquired with its Convex purchase.
Why Does HPE Need SGI?
The fact is, SGI is a serious player in the HPC market. It is a, if not the, primary competitor of HPE in this space. And the cost is a true bargain-basement price for the company. True, its margins have been negative—it lost almost $39M (net income), and its revenues have been on a downward spiral—but there is no denying its engineering credentials. In fact, in the announcement of the deal, HPE itself referenced SGI’s HPC and big data analytics credentials. According to executive VP and GM of the company’s enterprise group Antonio Neri, “At HPE, we are focused on empowering data-driven organizations. SGI’s innovative technologies and services, including its best-in-class big data analytics and [HPC] solutions, complement HPE’s proven data center solutions.” This is all well and good, but it is just the usual acquisition announcement love fest.
In the last fiscal year, SGI generated $532.9M in revenues, of which $138M came from its services division. This follows a slightly upward trend of the previous two years. However, a worrying dip in the current fourth quarter of fiscal 2016, when it would normally be up, no doubt made the SGI top brass give serious consideration to becoming part of a larger organization that can sell on a more global basis. Its services division, although not growing, is still an appreciable portion of the company. Of the revenue made this year, $394.8 million came from products and $138.1 million came from services. However, the gross margins for services are running at 42.7 percent for the year, compared to 21.7 percent for the products. This is precisely the kind of services business that HPE does not want to spin out, unlike its old EDS division, recently offloaded to CSC.
That said, SGI certainly needed to do something to lower its hardware costs and increase margins and revenues. However, it will now be able to take advantage of the HPE supply chain to significantly reduce those costs. It is expected that other savings will most likely be found in workforce reduction in areas where it makes sense, such as when back-office functions are eliminated post-merger. This will make the division leaner and stronger. SGI’s current worldwide staff count is 1,100.
It is not expected that the SGI merger will add anything to the HPE bottom line; it is expected to be neutral to its earnings. However, this is a very canny purchase at a bargain-basement cost for an underperforming company with excellent engineering staff and technology patents but no real path to success. It is probably the best outcome for most of SGI’s employees and its investors.