Cray’s president and chief executive, Peter Ungaro, took the opportunity of the opening of the company’s new European headquarters in Bristol, England, this week, to set out with crystal clarity his vision for Cray’s strategy and future direction.
Both in his remarks inaugurating the new HQ and in an interview afterwards, Ungaro was uncompromising that Cray’s interests are best served by providing technology that is clearly differentiated from the competition. It is interested in customers who want high-end computing and also, in this age of big data, in those who want advanced analytics. Those who have problems that could best be solved on commodity hardware should, in effect, look elsewhere.
The opening of the new EMEA (Europe, Middle East, and Africa) headquarters comes just two months after Cray announced that was appointing a dedicated vice president of sales for the region, Catalin Morosanu. Cray has had several recent successes in sales in Europe, with operations due to start in September at the UK Met Office, which had placed Cray’s biggest single contract outside of the USA; the start-up of the Hornet system at Stuttgart; and the announcement in March that Petroleum Geo-Services (PGS), a global oil-and-gas company headquartered in Oslo, Norway, is to buy a five-petaflop XC40 supercomputer together with a Cray Sonexion 2000 storage system, in what will be one of the largest supercomputers deployed in the commercial sector.
European expansion
But Ungaro was clear that there was a lot further to go in Europe. Many of Cray’s current customers -- and some whom it hopes may become customers in future – were special invitees to the official opening. Ungaro sees an opportunity not only for growth within the range of Cray’s existing customers – many of whom are Government or academic customers -- but also big opportunities in the commercial sector and especially in data analytics. Industry is going to be a big part of our growth in future,’ he said.
One part of Cray’s strategy in Europe, according to Ungaro, will be to partner with solutions’ providers. While Cray will provide the compute power, the solutions providers will be the ones offering software, training and other such services to make a complete solution for the customer. ‘A lot of these solutions will be in single vertical segments of the market,’ he explained. ‘Cray is a good partner to work with but we’re not a services company. We want to be the underlying technology provider and partner with the solutions providers.’
Because Cray is a relatively smaller company than, say, IBM, it was easier, he said, to stay really focused and also grow very fast. It was great place to be, he continued, pointing to the doubling of Cray’s revenues over the past few years.
Convergence of HPC and Big Data
The explosion of data was not slowing down and, in addition to Cray’s traditional core areas of modelling and simulation, analytics and data would be of growing importance to the company. He illustrated his points, both about the focus on high-end analytics and the importance of data, by recounting the story of a financial services company – which he was unable to name for reasons of commercial confidentiality. It has, what it believes to be, the largest private commercial cloud in the world.
But some 15 to 20 per cent of the company’s applications did not run well on the cloud. Even in preliminary discussions with Cray, it became apparent that the tightly integrated system that Cray provides would offer a significant return on investment when applied to these applications that were not running well on the more distributed cloud environment. ‘So we are looking to put a Cray supercomputer into the cloud to run just that 15 per cent of applications,’ Ungaro said.
In a further restatement of the clarity of the commercial vision that he had set out for Cray, Ungaro continued: ‘For Cray, just 15 per cent of those applications is super-exciting. We see a huge opportunity there.’
Ungaro repeatedly stressed his view that supercomputing and big and fast data are converging. In addition to the mathematical models which have been the traditional raw material for computation, increasingly people were building data models, he argued, and these ‘give us a bigger and richer view of the world’. Cray’s response is summed up in its theme of ‘compute, store, analyse’ and in the face of the convergence of HPC and big data, the company’s strategy was to ‘build a single system that can do it all. We see the opportunity to change the game by applying HPC rather than commodity computing. We are looking for customers who want to do advanced analytics that’s where we have an advantage over Oracle and IBM.’
It was, he said, a big step for the company to push harder into the commercial and big data sectors: ‘It’s a really large opportunity for us.’
