Session
Leading edge digital IP
Session Chair
DescriptionDigital IP design has to improve rapidly to catch up with the speed and performance requirements for massive compute intensive applications in the AI era. IP designers are exhausting every possible combination to improve power, performance and area efficiency of digital design. In this section, we talk about algorithmic changes that double the throughput of an Ethernet switch, a fast calibration method for PHYs, an FPGA based hardware accelerator, AI based technology library selection to optimize the PPA, arbiter logic for high performing systems and a highly efficient CXL memory compression design for data center applications.
Event TypeIP
TimeMonday, June 241:30pm - 3:00pm PDT
Location2012, 2nd Floor
Engineering Tracks
IP
Presentations
1:30pm - 1:45pm PDT | 100Gbps class in-vehicle Ethernet Switch architecture for next generation autonomous driving car Presenter | |
1:45pm - 2:00pm PDT | Hybrid Tiled Vector Systolic Architecture to | |
2:00pm - 2:15pm PDT | Evaluating power, performance, and area for standard cell libraries from different IP providers | |
2:15pm - 2:30pm PDT | Window Feedback Based Multi-Master Arbiter IP for Efficient Hardware Resource Sharing Presenter | |
2:30pm - 2:45pm PDT | An All-Digital IP for Fast Correction of Time-skew Mismatch in Time-Interleaved Analog to Digital Converters for Communication Receivers | |
2:45pm - 3:00pm PDT | Open Compute Platform(OCP) ready Hardware Accelerated CXL Memory Compression IP for Data Center Applications Presenter |