Machine learning and big data in Ansys simulation

Posted 9 January 2017
machine learning big data Ansys simulation

Ansys presents Ansys SeaScape, a software architecture combining elastic computing, big data and machine learning with the physics-based world of engineering simulation. Already available in Ansys SeaHawk, the product delivers accelerated design optimization of next-generation chips, for lower-power and high-performance electronics in mobile, data center and IoT markets.

SeaScape leverages big data technologies, such as elastic compute and MapReduce, to manage the tremendous amounts of data generated in engineering simulations, billions of variable in the case of a typical integrated circuit, helping supercomputing resources keeping the pace with the demand for higher fidelity simulations needed for increasingly complex products. The combination of big data techniques and Ansys’ capabilities in SeaHawk generates a broad range of capabilities to reduce size of the chip and its power consumption, without sacrificing performance and accelerating free design optimization and innovation without constraints. Early users of the architecture have realized an average of 5% reduction in die size and increased performances, using flexible low-memory compute to drive in-design optimization and sign-off simulation power, to benefit the earlier stages of the chip development cycle.

Contenuti correlati

Scopri le novità scelte per te x