After a rough first quarter, AMD executives dump the dense server market as it focuses on its server, PC, embedded and semi-custom chips.
U.S. regulators believe that the Tianhe-2 and Tianhe-1A supercomputers have been used in nuclear tests and go against national security interests.
The vendor also is growing its enterprise portfolio with two new networking switches and a low-cost interconnect module.
The DoE wants the vendors to build a system that is five to seven times more powerful than today's supercomputers in a push toward exascale computing.
Altera officials reportedly turn down a per-share price of more than $50, though at least one analyst doubts whether the deal is really off.
The vendor said faster SPARC64 X+ chips will help the M10-1 and M10-4 servers, jointly developed with Oracle, increase performance by 30 percent.
Demand is growing for a new generation of enterprise computing professionals armed with the expertise to address the mobile, big data and analytics trends. One way to prepare the next generation of computing pros is to train programmers to work with the mainframe. IBM's mainframes have long served as the core hub by major businesses for processing data and transactions. As mobile transactions are growing exponentially, the recently unveiled z13 takes that capability into the digital economy and helps companies meet customer expectations for speed and safety for trillions of transactions in the mobile economy. Learning to work with mainframes gives young programmers a career edge. Hosted at high schools, colleges and universities across the globe, the IBM Master the Mainframe contest is designed to give students mainframe knowledge and real-world experience using enterprise computing skills. 74,000 students across the globe have competed in the Master the Mainframe contests since they began in 2004 and 8,100 students competed this past year. The program will crown overall global winners next year with the 2016 World Championship. Take a look at this year's winners.
King's College London now shares data center facilities with five other institutions to pool high performance computing power.
Cray will bring a 5-petaflop system to oil-and-gas company PGS, while SGI will upgrade Total's ICE X Pangea supercomputer.
Supermicro will integrate its servers and storage appliances with Arista networking to create converged offerings for virtualization and big data.
The Knights Landing chip will offer 3 teraflops of performance and three times the single-thread performance of the current Knights Corner chip.
The server gear the consortium unveiled at the OpenPower Summit showed the speed of innovation that open development can bring, officials say.
SAN JOSE, Calif.—OpenPower Foundation officials used the group's first conference to show off more than a dozen new pieces of hardware from its members, to talk about how the consortium operates and try to recruit new members, and to position itself as an alternative to Intel in the data center. OpenPower was launched in December 2013 as a way of leveraging IBM's Power architecture to build products that address demands from hyperscale and Web-scale organizations for greater performance and efficiency. IBM has opened the Power architecture to allow the more than 110 OpenPower members to build products atop the platform. At the OpenPower Summit here—run in conjunction with the GPU Technology Conference 2015—foundation officials stressed that the open development business model enables faster innovation than Intel's way of doing the design, development and manufacturing work itself. "Without a community, we don't get innovation," said Randall Ross, a foundation official and Ubuntu community manager with Canonical. "We get something else. We get a monoculture." This eWEEK slide show takes a look at the hardware products introduced March 18.
SAN JOSE, Calif.—Nvidia CEO Jen-Hsun Huang is promising that his company will be a significant player in the growing field of deep-learning technology, where neural networks can learn from past experience and improve their performance over time. Deep learning has use in a broad array of areas, from self-driving automobiles to medical services to image and voice recognition. At the GPU Technology Conference 2015 (GTC) here, Huang introduced a new powerful GPU, the GeForce Titan X, as well as Digits DevBox, an appliance powered by the new GPU that runs Nvidia's Digits software to help developers and data scientists researching deep-learning technologies. The new offerings build on Nvidia's efforts over the past decade in expanding the use of its GPU accelerators in mainstream, scientific and high-performance computing (HPC). At the GTC, Nvidia showed off its new offerings, while other vendors, such as Dell, Hewlett-Packard and Lenovo, displayed their own compute systems that leverage Nvidia's GPU technologies. This eWEEK slide show gives a sampling of some of those systems as well as some of Nvidia's newest products.
At the first OpenPower Summit, the group announces more than 10 offerings from the likes of IBM, Tyan, Rackspace and Cirrascale.