Larry Ellison has become a “chip zealot” who talks more and more about SPARC and how to ramp up the hardware side of Oracle’s business.
So said Mark Hurd, co-president of Oracle, at an event celebrating 25 years since the launch of the first SPARC processor, a tech-star studded event held at the Computer History Museum in Mountain View, Calif.
Hurd promised that Oracle (with Ellison’s blessing) would seize the microprocessor lead from IBM over the coming year, thanks to its investment and commitment in the business. “We have lots of money. Because we have an owner that counts it,” he joked, drawing chuckles from the audience.
Since acquiring Sun back in 2009, Oracle has moved to make the firm’s hardware operations key to its systems strategy, combining it with Oracle firmware and databases to create fully integrated systems for big data analytics and the cloud.
“We’re doing more than just the silicon, we’re putting more software directly on the microprocessor,” said Hurd, promising those who had been with SPARC from the start that the technology was “in good hands.”
Sun Microsystems, born in 1982 as a tiny startup, launched its reduced instruction set computer (RISC) architecture back in 1987, when it debuted the Sun-4, the first SPARC-based computer.
“When [Sun Microsystems co-founder] Scott McNealy stood in front of a press conference 25 years ago and said SPARC would revolutionize the industry, few in the room, including me, took him seriously, but SPARC has now proven it has some staying power,” said Nathan Brookwood, research fellow at consulting firm Insights 64, who attended both the SPARC launch and the anniversary celebration 25 years later.
This, said Brookwood, was all the more amazing, given that Sun in 1987 lacked the industry clout of Digital Equipment with Alpha or Hewlett Packard with PA RISC, two architectures that have since come and gone.
“SPARC never achieved the broad multi-vendor adoption Sun had hoped to achieve, but has proven to be a durable design,” he said.
In a video aired at the event, McNealy himself told the audience how important he felt the SPARC platform was, from its inception.
In the context of a tiny startup competing against giants like Motorola, IBM and Intel, McNealy said it had been hard to win over the skeptics, but that nobody was delivering the performance or the technical computing power needed for the engineering workstation space. “So we had to do it ourselves,” he said.
Rival companies were too focused on the PC market, said McNealy, while Sun chased more horsepower, leading the team to the invention of its own architecture, RISC. It was, said McNealy, “simple computing and simple architecture,” though one that was far more scalable and powerful than competitor products at the time.
“We pretty much knew we had a hit when we eventually got the chip back, plugged it in and it started wiggling and we saw the benchmarks,” said McNealy, with pride.
At the time, he explained, the engineering watermark was the VAX (Virtual Address eXtension) instruction set architecture and its 1 MIP performance. “All of a sudden we were launching a little deskside computer server at 7 MIPs at the low end and 10 MIPs at the high end in a desktop workstation,” said McNealy, pointing out that it was also noticeably cheaper than VAX.
“It was pretty clear from a price performance standpoint that we had a huge winner,” he said.
IBM has the benefit of a captive foundry, so they can produce unique technology for their microprocessors. Oracle doesn't have that advantage... but then, Larry Ellison is one of the few people who has a bank account as large as his ambitions, so maybe things could get interesting.
Hurd was actually quite a funny guy at the presentation yesterday, but you're right that funny does not a good businessman make.
Oracle is no slouch of a company though, and it seems to be taking its acquisition of Sun seriously,so it could yield some promising results.
Larry Ellison is a legend in his own mind.
Talk about someone with a god complex. If Mark Hurd does for Oracle what he did for HP, then I think Oracle will become even more irrelevent to modern technology.
Just my opinion.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.