Cloud native EDA tools & pre-optimized hardware platforms
Co-authored by: Srinivas Kodiyalam, Senior Developer Relations Manager, Industrial HPC and AI at NVIDIA and
Samad Parekh, Senior Staff, Product Marketing Manager, Custom Design and Physical Verification Group at Synopsys
With the ever-increasing demand for more computing performance, the HPC industry is moving towards a heterogeneous computing model, where GPUs and CPUs work together to perform general-purpose computing tasks. In this heterogeneous computing model, the GPU serves as an accelerator to the CPU, to offload the CPU and to increase computational efficiency. To exploit this computing model and the massively parallel GPU architecture, the application software will need to be redesigned. Synopsys and NVIDIA engineers have been working together to use GPUs to accelerate circuit simulation technology.
IC Design complexity has continued to grow exponentially. Just in the last decade, as process technology has advanced from planar to FinFET technology, device counts and parasitics have increased significantly. Modern process technologies also require verification across many more process, voltage, and temperature corners, resulting in design complexity increasing by orders of magnitude. During this time CPU performance gains have largely plateaued while GPU performance has been growing and continues to scale well beyond Moore’s Law. These trends will only further increase the gap between the two computing methodologies over time. GPUs provide significant advantages in dealing with circuit simulation challenges including the high volume of floating-point operations and memory bandwidth usage as well as the large number of massively independent computing jobs with large post-layout circuits.
Synopsys announced today at its SNUG World international user conference PrimeSim™ Continuum, a unified workflow of best-in-class SPICE and FastSPICE technologies, speeding the creation of analog, RF, mixed-signal, custom digital and memory designs. The PrimeSim Continuum solution is built on innovative SPICE and FastSPICE architectures providing design teams 10X faster simulation while maintaining signoff accuracy. PrimeSim’s next-generation architecture with unique GPU technology delivers significant performance improvements needed to perform comprehensive analog and RF design analysis while meeting signoff accuracy requirements. Benchmarks models run on NVIDIA DGX systems show speedups ranging from 4-12X over multi-core CPUs. While performance gains are across the board on a variety of circuit types, the best improvements are seen when running large post-layout simulations. When coupled with long transient run times, the performance improvement is even more noticeable.
PrimeSim achieves its most impressive performance gains by leveraging the massive parallelism of the CUDA GPUs. The core technologies involved are:
Nano-scale IC simulation complexity and exploding model sizes necessitate the need for multiple GPUs with extremely fast interconnections. The NVIDIA DGX systems and HGX™ platform bring together the full power of NVIDIA GPUs, NVIDIA® NVLink®, NVIDIA Mellanox® InfiniBand® networking, and a fully optimized NVIDIA AI and HPC software stack to provide the highest application performance.
Learn more about PrimeSim on GPU at the SNUG World 2021 user conference, running from April 20-22. NVIDIA will participate in SNUG 2021 both in a panel: “The Future of Simulation Hardware” and present in the Custom-AMS track titled, "Update on Hardware Computing Technology and its Impact on Simulation.” Register to see all the sessions.