|
||||||||||
Wave Semiconductor Licenses SonicsGN On-Chip NetworkSonics’ NoC Provides Superior IP Integration, Multi-channel Memory Access, and SoC Performance and Power Management Capabilities Milpitas, Calif. – October 13, 2015 – Sonics, Inc., the world’s foremost supplier of on-chip network (NoC) technologies and services, today announced that Wave Semiconductor (www.wavesemi.com) has licensed SonicsGN®, the company’s flagship NoC product. Wave is using SonicsGN to integrate a massively parallel array of custom processing elements with third-party IP in its system-on-chip (SoC) design for machine learning and other compute-intensive applications. “Wave Semiconductor is delivering on the promise of reconfigurable computing systems that fundamentally changes the way big data is processed and analyzed,” said Grant Pierce, CEO of Sonics. “SonicsGN is particularly well suited to Wave’s SoC design because it is the only NoC on the market that includes patented, interleaved multi-channel technology (IMT) to efficiently use the massive bandwidth available from the latest memory subsystems.” “We selected the Sonics NoC because it offered us the features and flexibility to meet our system throughput and power requirements, while easily adapting to our high performance and challenging floorplan requirements,” said Chris Nicol, CTO of Wave Semiconductor. “Our programmable computing fabric has a unique interface and requires huge bandwidth to support the multiple memory ports on our SoC. Sonics’ NoC provided the memory access capabilities we specified and the interconnect technology fits well with our blend of third-party IP and full-custom proprietary circuitry.” For more information on Sonics’ customer-proven and patented IMT, download the white paper. About Wave Semiconductor Wave’s programmable technology is based on massive arrays of 12GHz byte processors and 32-bit MAC units – called a Byte-Fabric™. Byte-Fabrics are real-time reprogrammable, finally delivering on the promise of reconfigurable hardware. This reconfigurable hardware uniquely allows Wave to move algorithms to the data rather than vice-versa, dramatically reducing power and increasing performance. By placing special emphasis on low-power computing, Wave offers best-in-class performance as measured in operations/Watt. About Sonics, Inc.
|
Home | Feedback | Register | Site Map |
All material on this site Copyright © 2017 Design And Reuse S.A. All rights reserved. |