Abstract
□ 3D integration: - Many opportunities to add system value, many interesting research and development challenges - An interdisciplinary view and approach - technology, VLSI, architecture - is essential □ 3D and 2D are complementary - The best solutions will leverage both appropriately, over time □ In the near term, need to introduce 3D with minimal disruption - Leverage the large, existing base of 2D design automation tools, methodologies, and designs - Add value to conventional system structures □ In the longer term, need to fully optimize for 3D - Different system balances between bandwidth, cache capacity, and computation - A roadmap that enables the software stack to evolve - What new systems will 3D integration enable? © 2010 IBM Corporation.