Your opinion about gps4us News:
Welcome to gps4us News About gps4us  |  News  |  Support & Contacts
Twitter Facebook Linkedin YouTube
Recommend this page
Bookmark and Share

Smallest Microchips and fastest Supercomputers empower next generation of Cloud Computing

by Russel 2011-11-07 11:14

The popularity of the voice recognition, audio and video-rich touch-sensing peripherals taking the market of personal mobile assistant, personal communicator and portable GPS navigating gadgets by storm demand for smallest, lowest cost tiny form factor micro processing platforms on the user end, and for the faster supercomputers on the Cloud Computing services back end. Answering the new market needs the technology innovators introduce more efficient computational platforms in micro and macro world of computing.

super computing microchipMicrochip introduces smallest, low cost feature-packed 32-bit micro controllers with interface for audio playback applications, capacitive touch peripherals USB 2.0 and digital pin remapping. A new series of low pin count 32-bit micro controllers provide over sixty dhrystone synthetic million operations per second of performance. The tiny supercomputer is manufactured in a small 5 mm x 5 mm form factor ideally suited for mobile user space-constrained and cost-sensitive handheld designs. The micro-chip features dedicated audio and capacitive-sensing peripherals along with a host of additional useful features that make this 32-bit micro controller suitable for mobile applications in the global positioning systems, consumer, industrial, medical and automotive markets relying on the mobile user portable personal assistant devices.

On the other rend of the Cloud Computing spectrum a Petaflop capable supercomputer known as the highest speed processing platform. The super Cloud Computing machine is currently becoming available for complex simulations to help solve complicated and difficult problems such as environment, energy and food data processing. The supercomputer delivers record performance for artificial intelligence, aeronautics, weather, intense three dimensional topographic and multi dimensional medical areas of computing including drug development and manufacturing.

Super fast PRIME HPC FX10 Petaflop supercomputerOn the other end of supercomputing Fujitsu Launches New FX10 technology adapting the best of the K-technology supercomputer. The new supercomputer further improves on predecessor technology jointly developed by the company and state-backed major research institute Riken. Early version of the K-technology supercomputer was ranked in June as the world's number one in current computing speeds. Building on the success the PRIME HPC FX10 is capable of quantum computing and theoretical processing performance of up to 23,200 trillion computations per second. Available early next year this super computing architecture named "HPC PRIME FX10" aggregates 98,304 nodes of IXfx 16 core SPARC64 processors. The enormous amount of computing power is spread over 1,024 racks and operates with  memory capacity of 6 Peta-bytes. As a result the HPC PRIME FX10 delivers the processing power of 23.2 Petaflops. This efficiency is more than twice that obtained recently by far most powerful supercomputer to date, the original "K" Computer, which last week managed to overcome the barrier of 10 Petaflops, or 10 quadrillion operations a second. In fact the record has been pushed pass 10 Petaflops and halfway toward 11, with peak at 10.51.

quantum supercomputing modelThe record speed of processing delivered by the first HPC PRIME FX10 has proven efficiency of the 16-core processor enabled IXfx SPARC64 architecture. High results come with price of extreme amounts of heat to take away from the breaking record system. To 23.2 Petaflops of processing capability required to use liquid cooling system that significantly improved system efficiency, in comparison to the previous K-technology super computing platform which was consuming about 9.89 megawatts per Petaflop, spanning 864 server racks that manage over 88,000 CPU-s inter-connected by high speed backbone matrix. The new super computing machine can run thousands of parallel real time simulators and other intense applications. 

Obvious success of the new architecture put HPC PRIME FX10 on top of the list of world's fastest supercomputing platforms ready for implementation as fully commercial model. The commercial version of the 'K' Technology supercomputer aimed at companies and research institutes requiring high-level computing capabilities is planned to reach the market early next year. The commercial model of the HPC PRIME FX10 will face the new challenges of modern society, such as the development of new drugs, prevention and mitigation and other measures as well as all kinds simulations, quantum computing and the analysis of earthquakes, weather events, tornado predictions, ocean research, deep space voyages computations. Originated in Japan the super platform will be marketed globally. With a price of about $650,000 for a single-rack model, the manufacturer can scale the system up to a 1,024 rack configuration. Over the next three years about 50 systems will be delivered to a range of groups of major customers including universities, research institutions and private companies. The customer base is widely using complex simulations in a variety of studies including artificial intelligence research, analysis of earthquakes and weather phenomena.


Comments are closed
Thank you for being gps4us News Reader

13431 Beach Ave. Marina Del Rey, CA 90292  | Customer Services | Contact Info & Support
About gps4us | Testimonials | News | Terms & Conditions | Privacy Policy | Shipping & RMA
© 2009-2011 GPS4US Online Store, All Rights Reserved.