Japanese tech giant Fujitsu has confirmed the demise of its mainframe business in the year 2035 and hinted it’s working on significant defense projects.
CEO Takahito Tokita confirmed the 2035 death date earlier this week when delivering a briefing on the company’s medium-term management plan and future directions. He used that moment to note that Fujitsu will turn 100 in 2035, share his view that the company will enjoy its most impactful years between now and its centenary, and suggest AI is the reason the company will remain relevant.
By 2035, he expects Fujitsu’s hardware engine will be either what he called “AI supercomputers” – powered by the “Monaka” CPUs it’s building with Broadcom and inferencing chips it’s working on with French concern Scaleway – or quantum computers. He said those machines will become mainstream workhorses in the same year as Fujitsu’s mainframes die,
The CEO said everything Fujitsu does in the future will involve AI, and the company will eat its own AI dogfood.
“We have built a globally standardized data platform,” Tokita said. “Starting this fiscal year, based on this data platform, we will accelerate the full-scale implementation of AI-driven management using our own AI. This will enhance both the speed and quality of decision-making and management judgments.”
Fujitsu will use its experience to make itself into a case study of clankers running companies.
The company will also persist with the “Uvance” brand under which it blends consultancy services and IT-as-a-service.
But the company will shift away from systems integration work and charging by the hour to what Tokita described as “an earnings structure based on value and outcomes.”
“We have long managed our business under a highly risky structure in which revenue was heavily concentrated in the fourth quarter,” the CEO told investors in a Q&A session. “We believe that achieving a more even distribution of revenue across quarters is extremely important for improving the quality of our management, although it is a gradual process,” he added. Customers are apparently on board with this plan, even though the CEO admitted “There is still plenty of room for further improvement in our pricing model, which outlines our approach for how we provide various technologies.”
More changes are in the works: “I would like to consider an approach in which we charge customers based on factors such as the workload of our personnel and the amount of data required to provide such technologies,” Tokita said.
Speaking of personnel, Tokita said Fujitsu Japan has stopped its annual graduate intake, and is instead now hiring people with specific skills it needs.
Asked about the importance of work for the defense sector, Tokita offered the following tantalizing response:
Fujitsu needs this plan to work, because full-year revenue fell 1.3 percent year-over-year to $22.3 billion, although profit popped 31 percent to $2.2 billion. ®