IBM reclaims world's quickest supercomputer title with Sequoia

You don't have permission to edit metadata of this video.

Edit media

Dimensions x Small Medium Large Custom

Subject (required) Brief Description Tags (separated by comma) Video visibility in search results Visible Hidden

Parent content

IBM reclaims world's quickest supercomputer title with Sequoia

Poster

Upload Preview

Ok

Cancel

The world has a new fastest computer after a team from technology giant IBM managed to create a new system to top the list. Sequoia achieved speeds of 16.32 petaflops per second in recent tests, therefore beating Fujitsu's K Computer which held the previous record with 10.51 petaflops per second.

According to BBC News, the petaflops scoring system is known as the Linpack Benchmark and this calculates the number of quadrillion floating-point operations per second a computer can do. In simpler terms, IBM explained that the computer could work out as many calculations in an hour as it would take 6.7 billion people using hand calculators 320 years to finish.

This, explained National Nuclear Security Administration administrator Thomas D'Agostino, will make it a welcome addition to a US nuclear research program.

"While Sequoia may be the fastest, the underlying computing capabilities it provides give us increased confidence in the nation's nuclear deterrent," he explained. "Sequoia also represents continued American leadership in high performance computing."

This success helps the US take back top spot in the list of supercomputers after China took it from the country two years ago. Furthermore, the US now has three computers in the top ten. Germany and China have two each, with France, Japan and Italy also having one each in the list.

Other statistics further reveal how impressive the Sequoia machine is. The computer features 96 racks that include 98,304 compute nodes and 1.6 million processor cores. Additionally, it consumes only 7,890 kilowatts of power on full load compared to K Computer's 12,659 kilowatts.

Around 4,500 square feet of floor space is also given up for 1.6 petabytes of RAM and, according to reports, the device was tested for 23 hours without a single core failing. LLNL division leader Kim Cupps told Ars Technica how significant an achievement this is.

"For a machine with 1.6 million cores to run for over 23 hours six weeks after the last rack arrived on our floor is nothing short of amazing," she explained. The device will now be used at the US Department of Energy's Lawrence Livermore National Laboratory in California, undertaking simulations on how to extend the life of nuclear weapons.