ElasticComputeCloud
Latest
Amazon stores 1,700 human genomes in the cloud
Not content with speeding up web browsing and hosting federal data, Amazon Web Services are now helping in the fight against disease. Bezos' crew is donating a chunk of free cloud storage to the 1000 Genomes project, which aims to make it easier for scientists to search for genetic variations linked to diseases. These gene-hunters can also use Amazon's Elastic Cloud Compute service to analyze data and discover patterns, although those functions won't come gratis. The DNA sequences of 1,700 mostly anonymous Homo sapiens from around the world have already been logged, but the project has to upload another 1,000 samples before it meets statistical requirements. If it ever needs fresh volunteers, perhaps a free USB gene sequencer and a Prime subscription might do the trick?
Behind Amazon's Silk browser lurks a really fast supercomputer
We were hardly shocked to see Fujitsu atop the most recent list of the world's fastest supercomputers, but perhaps more surprising is the fact that Amazon cracked the top 50, as well. Turns out, the company's Elastic Compute Cloud (EC2) servers are powered by a Linux-based, 240-teraflop beast that boasts 17,024 cores, 66,000 GB of memory, and a ten gigabit Ethernet interconnect. That's good for 42nd place on Top 500's global rankings, and it's also good enough to power Silk, the browser you'll find on the Kindle Fire. But Amazon has a long way to go before catching up with the Fujitsu K, which recently cracked that vaunted ten petaflop barrier.
Amazon's Elastic Compute Cloud on-demand computing service
Amazon continues to push the boundaries of web services and people's tolerance for clever branding with its recently launched "Elastic Compute Cloud", offering easily scalable computing on demand to developers. The service works in conjunction with Amazon's Simple Storage Service (otherwise known as S3), giving developers a virtual computing environment within which they can do what they please. Each virtual server instance they get gives 'em the equivalent of a 1.7Ghz Xeon CPU, with 1.75GB of RAM, 160GB of disk space, and 250Mb/s of network bandwidth, of which they can of course get as many as they like (or can afford), with prices starting at just $0.10 per instance-hour consumed and $0.20 per GB of data transferred outside of Amazon. Sure seems cheap compared to what it'd cost to get up and running on your own, although it'll obviously still add up pretty quickly if you're working on large scale projects. [Via GigaOM]
