Ansys and TSMC have teamed up with Microsoft to enhance the simulation and analysis of silicon photonic components.
The companies recently revealed that, by using Microsoft Azure NC A100v4-series virtual machines, they were able to speed up the Ansys Lumerical FDTD photonics simulation by more than 10x.
Silicon Photonic Integrated Circuit (PIC) technology is a type of optical communication that speeds up data transfer speeds and is used in data centers and IoT applications.
Because the process of combining photonic and electronic circuits is very challenging, and even small mistakes can be expensive and result in time setbacks of up to several months, simulation and analyses are conducted to assist the procedure.
Ansys conducted its simulations using Microsoft Azure virtual machines based on Nvidia A100 GPUs, which the company claims has enabled it to identify optimal chip designs as much as 10x faster.
“The size and complexity of our multiphysics silicon solutions makes the process of simulating all possible parameter combinations challenging,” said Stefan Rusu, head of silicon photonics system design at TSMC. “This latest collaboration again highlights that Ansys effectively harnesses the latest cloud infrastructure and techniques to deliver powerful, predictively accurate solutions that produce results in a fraction of the time.”
Shelly Blackburn, CVP of Azure infrastructure, digital and app innovation at Microsoft, added: “Our collaboration is a significant advantage for users seeking the combined power of HPC and AI (artificial intelligence), using the flexibility of cloud solutions while maintaining the familiar on-premises experience.
“By working together, we aim to address the complexities of large-scale designs essential for high-quality semiconductor products. Utilizing the power and scalability of Microsoft Azure’s cloud computing is a key strategy in overcoming these challenges.”
The Microsoft Azure NC A100v4-series virtual machines were launched in August 2024. The series is powered by Nvidia A100 PCIe GPU and third-generation AMD Epyc 7V13 (Milan) processors. The VMs feature up to four Nvidia A100 PCIe GPUs with 80 GB memory each, up to 96 non-multithreaded AMD Epyc Milan processor cores, and 880 GiB (Gibibyte) of system memory. According to Microsoft, the VMs are ideal for applied AI workloads.
Ansys and TSMC also recently announced that they had expanded their collaboration to use AI for 3D-IC design and develop solutions for advanced semiconductor technologies for data center, high-performance computing, AI, and wireless telecommunication applications.
Earlier this month, Ansys inaugurated a high-performance computing data center in Pune, India. The software company now has three on-premises data centers, the newest of which will be used to support its R&D operations and increase its capability to serve customers across sectors, including automotive, aerospace, and healthcare.
Ansys, an engineering and product design software firm headquartered in Pennsylvania, was acquired by Synopsys in January of this year. The UK's CMA opened an investigation into the acquisition in August.
As of this week, TSMC is reportedly planning to build more chip fabs in Europe with a focus on AI semiconductors.
