Thalia Design Automation in Wales has boosted its analog IP tool with a new machine learning algorithm.
The improved ML algorithms in Thalia Design Enabler help analog designers balance circuit and layout requirements, factoring in the impact of layout changes. This helps simplify the migration of analog IP to new process technology nodes by preserving, improving and validating the topology and the floorplan.
- Thalia takes machine learning in EDA tool down to 22nm FDSOI
- New CEO for Thalia Design Automatio
- NXP’s Catena Group selects Thalia for analog reuse
Design Enabler is part of the powerful AMALIA IP Reuse Platform that enables re-use of analog and mixed signal integrated circuits. By quickly identifying the minimum set of devices needed to adjust the design the tool reduces the number of changes required in layout while improving performance. The enhanced ML algorithms can make savings of up to 50% in time and resources.
Thalia has now deployed Machine Learning (ML) across all of its AMALIA platform to analyze waveforms and to propose the ideal solution for the designer, based on differential analyses.
“Customers using Design Enabler tell us they’re saving at least half of their original re-design time, and with the enhancements we’re announcing today, we expect our customers to see immediate benefits both in optimizing the performance of their design and quickly and easily being able to balance this against the best layout to suite their design,” said Christopher Yates, VP Software Engineering at Thalia.
Other articles on eeNews Europe
- Ultra-thin electrode for 1nm atomic transistors
- US fights back with first exascale supercomputer
- Infineon plans to drastically increase production investments
- SiPearl confirms Nvidia supercomputer deal
- ARM battles RISC-V at Renesas
- imec charts path beyond 1nm