Nvidia's latest accelerator is also featuring NVIDIA GPU Boost technology, which converts power headroom into a user-controlled performance boost.
It is based on the NVIDIA Kepler compute architecture and offers 4.29 teraflops single-precision and 1.43 teraflops double-precision peak floating point performance, according to Nvidia.
The Tesla K40 GPU also includes 12GB of GDDR5 memory clocked at 3GHz; 2,880 CUDA parallel processing cores; Dynamic Parallelism and; a PCIe Gen-3 interconnect support.
Nvidia did to raise the thermal threshold for the new card as both the K20X and K40 are 235W solutions. The also K40's base core clock has been also increased to 745 MHz, with boost clocks reaching the 810 and 875 MHz:
The Texas Advanced Computing Center (TACC) at The University of Texas at Austin plans to deploy "Maverick," a new interactive, remote visualization and data analysis system powered by NVIDIA Tesla K40 GPU accelerators. Maverick is expected to be fully operational in January 2014.
Nvidia says that more than 240 software applications take advantage of GPU acceleration.
The NVIDIA Tesla K40 GPU accelerator is available now and in the coming months from a variety of server manufacturers, including Appro, ASUS, Bull, Cray, Dell, Eurotech, HP, IBM, Inspur, SGI, Sugon, Supermicro and Tyan, as well as from NVIDIA reseller partners.
Nvidia wants to bring its GPU accelerator technology beyond the realm of supercomputing and into the heart of enterprise-scale data centers. The company today said that it plans to collaborate with IBM on GPU-accelerated versions of IBM's portfolio of enterprise software applications on IBM Power Systems.
NVIDIA and IBM also plan to integrate the joint-processing capabilities of NVIDIA Tesla GPUs with IBM POWER processors.
The partnership between NVIDIA and IBM builds on the OpenPOWER Consortium, in which IBM, NVIDIA, Google, Mellanox and Tyan aim to establish an open ecosystem based on IBM's POWER architecture.