Bitcoin network reaches size half that of Tianhe-1A, quarter of Folding@Home
Bitcoin ranks at about a quarter as powerful as Folding@Home — the most powerful distributed computing cluster in the world.
The image in this post is created in real time, however the metrics are bing compared from the February 25th, 2011 levels. Information gathered from Raulo’s comments on the forum.
US reclaims the crown with the world's fastest supercomputer
Suck it, China and Japan… with IBM’s Sequoia supercomputer, we have once again regained the title of world’s fastest supercomputer. Merka.
Japan re-claims the title of having the world's fastest supercomputer. Fine, whatever, see if I care.
A couple months ago, China proudly unveiled what was at the time, the world’s fastest supercomputer that was clocked at 2.5 petaflops. Now, Japan’s K Computer has absolutely smashed that record, computing 8.2 quadrillion calculations per second, or 8.2 petaflops. EAT IT COMMIES.
IBM wants to build an exascale computer by 2024
IBM and ASTRON have embarked on a huge project with the goal of building an Exascale computer by 2024. Said computer would collect data from a 3000 km array of radio telescopes, streaming an exabyte of data every day. How much data is that? That’s twice as much data as is in the entire internet in a single day. So imagine all the data going around the internet on an average day around the world and double that. That’s how much data is going to be collected and processed by these radio telescopes.
“SDSC’s project is another example of cloud computing expanding the accessibility of high-performance computing (HPC) functionality once reserved for an exclusive set of institutions. Instead of being forced to build out huge clusters inside your own data centers, customers can outsource supercomputing needs to cloud vendors. Amazon offers special cluster compute instances for just such a purpose, and even built a supercomputer on the Elastic Compute Cloud that ranked among the Top 500 supercomputing sites in the world. Another project recently featured by Ars used the Amazon compute cloud to build a 30,000-core cluster for a pharmaceutical company that ran for about seven hours at a peak cost of $1,279 per hour.”—
Sequoia Throws Petaflops at Fusion Simulation
by Txchnologist Staff
One of the world’s fastest supercomputers has performed a record number of simulations to help physicists in their quest to produce fusion energy.
Lawrence Livermore National Laboratory researchers set all of their Sequoia supercomputer’s 1,572,864 processors (known as cores) on a single problem: modeling the motion of charged particles after a powerful laser beam strikes a dense cloud of ionized gas, called plasma. Using all of its cores, Sequoia can process 16.3 quadrillion calculations per second.
The machine is letting the team follow the simultaneous evolution of tens of billions to trillions of individual particles and the electromagnetic interactions between them. Such simulations are used extensively in plasma physics.