JaguarRoute

  Our Technology

With JRoute’s cutting edge technology, the traditional sequential routing flow consisting of global routing, pin assignment and detailed routing is discarded. Instead, these three sub modules are integrated seamlessly into a single module and executed simultaneously to avoid back-and-forth iterations and eliminate the inaccurate estimations that lead to slow routing.

Using JRoute, a design is partitioned first and then routed independently. The conflicts between partitions are resolved before stitching them together into the full-chip design. Based on such an approach, the routing technology becomes not only scalable, but also suitable for utilizing multi-core CPUs to boost the speed further. This feature will increase JRoute’s speed tremendously in the next phase of development.

Because of JRoute’s superior capacity, the user does not need to perform the conventional block routing first and then chip assembly. If a placer can place the entire design without partitioning it into sub-blocks, JRoute will be able to route the complete design. The capacity of JRoute is simply limited by the size of RAM.

JRoute is designed to follow the grid first. If the grid approach is not successful, a shape-based engine is used. JRoute is classified as a dual-mode(grid-based and shape-based) router.