A new ranking out this morning shows that an IBM system at Lawrence Livermore National Laboratory has broken the world speed record. The IBM Sequoia can do more than 16 quadrillion floating-point calculations per second, or 16 petaflops. That blows away the previous leader, a Japanese system, that does 10.5 petaflops. Lawrence Livermore uses its ultrafast computers for complex simulations, and for developing code in energy, life science and military applications.
Ok, so what’s a petaflop? FLOP stands for “floating-point operations per second.” And peta sits between tera and exa. But how fast is that in plain English? Think about how you measure hard drives. We know that tera is bigger than giga, which is bigger mega. Personal computers today are clocking in the mega range.
This Forbes article puts the Sequoia in perspective by comparing its 1,572,864 CPU cores with the 4 in the computer he was using to write the story.
Lawrence Livermore is no newbie in the high-performance computing game. In a story looking at the history and future of these supercomputers, KQED’s Quest learned about the Cray-1 (right).
Supercomputers have improved at a break-neck speed, especially if you look back to the Cray-1. In 1976, this six-foot tall tower of wires was the most powerful supercomputer the world had ever seen. It was installed at Lawrence Livermore National Lab for fusion research.
“If you needed an icon for a supercomputer, you would use the Cray-1,” says Dag Spicer, senior curator at the Computer History Museum, where the computer is spending its retirement. “It blew people’s minds. It was so powerful, so fast.”
Of course, in today’s terms, “It’s roughly equivalent to a first generation iPhone from Apple,” says Spicer.
And what operating system is ‘Sequoia’ running, you ask? Linux. Who says penguins aren’t fast?