Nvidia Upgrades GPU For High-Performance Computing
The Tesla 10P is for companies that need high computational power to drive apps found in scientific research, oil and gas, financial, and medical industries.
Nvidia's Tesla C1060 card slips into a PCIe slot for delivering high-performance computing to workstations. The card delivers 1 teraflop of power for $1,699.
Nvidia on Monday launched a new version of its general-purpose processor for companies that need high computational power to drive applications found in scientific research, oil and gas, financial, and medical industries.
The Tesla 10P is the company's second generation, general purpose graphic processing unit for high-performance computing. The latest product has twice the performance of the previous generation Tesla 8 product, or 1 teraflop of computational power versus 500 gigaflops. The T10P also has more than double the memory at 4GB versus 1.5GB in the older model.
With Tesla, Nvidia gives companies the option of adding a general purpose GPU to their server rack or workstation as an alternative to adding more dual-core or quad-core processors from Intel or Advanced Micro Devices. Nvidia claims its latest processor with 1.4 billion transistors and 240 processing cores delivers 10 times the computational power as a computer with two quad-core CPUs while using about the same amount of energy.
Nvidia, best known for selling graphics processors to drive top-of-the-line PC videogames and in workstations used by graphics professionals, offers the latest Tesla processor in two form factors. The first is in a 1u system, called the Tesla S1070, which fits in a standard 19-inch server rack. The other, called the C1060, is a workstation card that slips into the PCI Express slot.
To use Tesla, developers have to learn Nvidia's CUDA technology. CUDA, or compute unified device architecture, allows programmers to use the C programming language to code algorithms for execution on the GPU. In high-performance computing, that part of an application that's heavily computational would be compiled using Nvidia's CUDA compiler in order to run on the Tesla-based server or workstation. Nvidia claims CUDA is not difficult for developers because it uses C, a language they are already familiar with.
The S1070 server and C1060 card are available now for $8,000 and $1,699, respectively. Each server comes with four GPUs, delving 4 teraflops of power per system. A teraflop is a trillion floating point operations per second. Nvidia first launched Tesla a year ago.
Server Market SplitsvilleJust because the server market's in the doldrums doesn't mean innovation has ceased. Far from it -- server technology is enjoying the biggest renaissance since the dawn of x86 systems. But the primary driver is now service providers, not enterprises.
Join us for a roundup of the top stories on InformationWeek.com for the week of December 14, 2014. Be here for the show and for the incredible Friday Afternoon Conversation that runs beside the program.