PARIS – Magma Design Automation Inc. announced it has introduced Talus Vortex FX, an IC implementation solution to fully utilize distributed computing for place and route.
In a discussion with EDA DesignLine, Mark Richards, Senior Technical Marketing Manager, Design Implementation Business Unit, Magma Design Automation, said Talus Vortex FX should be viewed as augmenting the company's place and route offerings by leveraging the underlying Talus Vortex 1.2 technology to provide best-in class capacity, throughput and results.
Talus Vortex FX, Richards continued, is the industry’s only Distributed Place and Route Implementation Tool that leverages the Talus Vortex technology in each of the distributed partitions to offer unrivalled throughput and capacity. "Managing, synchronizing and ensuring optimum QoR for these partitions and the design as a whole is the overriding technology that marks the tool out against the rest of the P&R space."
Talus Vortex FX is claimed to be up to three times faster than Magma's new Talus 1.2. It provides higher capacity, enabling designers to implement multimillion-cell designs flat and to generate 2 million to 5 million cells per day with crosstalk avoidance, advanced on-chip variation (AOCV) and multi-mode multi-corner (MMMC) analysis enabled – while using existing hardware resources. Talus Vortex FX is suitable for very large, complex advanced-node designs.
Commenting on Talus Vortex FX performances versus Talus 1.2, Richards declared: "The throughput from design to design will always vary with any number of factors including things as simple as very high placement utilization to more esoteric reasons like complex MVDD requirements or highly channeled designs with difficult to meet topology and timing constraints. But across a broad number of real world designs that Talus Vortex 1.2 has been benchmarked on, we see consistent throughput at over one million cells per day."
Talus Vortex FX builds on the capabilities of Talus 1.2 to deliver enhanced capacity and "up to three times additional throughput". Talus Vortex FX integrates the Distributed Smart Sync technology to enable concurrent implementation management and smart synchronization of very large designs across a network of machines.
Asked for precisions on the Distributed Smart Sync Technology, Richards said it is aimed at optimally partitioning the design across the available (automatic or manually driven) resources to ensure maximum throughput whilst delivering best in class QoR. All of the normally complex tasks related to timing and physical partitioning is automatically handled by the tool.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.