The new machine learning technology combined with the core computational software allows Xcelium ML to deliver up to 5X faster verification closure on randomized regressions.
Xcelium ML is able to learn iteratively over a whole simulation regression through the use of computational software and the proprietary machine learning technology that is directly interfaced with the simulation kernel. The tool analyzes patterns that are hidden in the verification environment and guides the Xcelium randomization kernel on the following regression runs to provide matching coverage with reduced simulation cycles.
Cadence’s Xcelium Logic Simulator delivers the core engine performance for SystemVerilog, VHDL, mixed-signal, low power, and x-propagation. It can support single-core and multi-core simulation, incremental and parallel build, and save/restart with dynamic test reload.
“Kioxia has effectively utilized Xcelium simulation for a variety of our designs, and it addresses our ever-growing verification needs,” said Kazunari Horikawa, senior manager, Design Technology Innovation Division at Kioxia Corporation. “With the new Xcelium ML, we’ve seen a 4X shorter turnaround time in our fully random regression runs to reach 99% function coverage of original, and plan to use this technology in production designs to shorten the time to market for Kioxia’s business.”
“Xcelium ML is a powerful technology and a great example of the significant opportunity we have to leverage machine learning in verification,” said Paul Cunningham, corporate vice president and general manager of the System & Verification Group at Cadence. “Logic simulation continues to be the workhorse of digital verification, and we are investing heavily in fundamental performance optimizations like Xcelium ML to deliver the highest verification throughput to customers using our flow.”
Xcelium ML is part of the Cadence Verification Suite and supports the company’s Intelligent System Design strategy, enabling pervasive intelligence and faster design closure.