HPE announced last month that it intends to acquire Cray Computers for $1.3 billion. As when adding SGI in 2016 HPE will not add much server revenue through the move (see my Figure).
Super- and other scientific computers require specialist suppliers; in comparison with the commercial market sales lead times can be very long, budgets small and often dependent on government grants and funding.
Over the year to the end of March 2019 the majority of Cray’s revenues were derived from Supercomputers, although it also made money from maintenance and support, storage and data management ad engineering services (see my Figure opposite).
Despite being a very different market for decades, the style of scientific computing has been becoming much more relevant to commercial enterprises over the last decade. Clusters, high-performance graphics processing and the quest to run complex algorithms in the fastest time possible are now an essential feature for many in Financial Services, Manufacturing (including the development of autonomous vehicles) and Marketing.
Adding Cray will give HPE new IP, patents and experts who will enhance its technical capabilities and help to close the gap with companies like IBM, Oracle and Huawei, which spend proportionately more on R&D.
Unlike with SGI I expect that HPE will leave much of Cray’s business intact after the acquisition gets the go-ahead, not least because the US Federal government is a major customer.
Cray’s international revenues have taken a downturn in the last few quarters (see my Figure opposite for the rolling 4 quarter totals) and I doubt they will improve for as long as president Trump pursues a policy of trade war and new nationalism reigns.
©ITCandor Limited – unauthorised copying of this content is illegal and will be rigorously defended by us through court action