講題:Smaller and Faster Android: Optimization and Toolchain Perspective
講者:廖世偉 Shih-wei Liao ( Tech Lead, Google )
自介:Shih-wei Liao is a tech lead at Google Mountain View while being on-leave from National Taiwan University. He received his BS from NTU in 1991 and his MS and PhD from Stanford University in 1995 and 2000, respectively. He has many years of product experience at Intel and Google and more than two dozens of patent applications. On the other hand, he also publishes extensively on multicores, compilers, programming systems, and blocking & affine partitioning theory which is included in the Dragon Book. (Dragon book is a bible in the field of compilers.)
摘要:How to systematically achieve smaller Android footprint and optimized Android execution? I'll introduce the design of the Android toolchain and present several case studies on optimizations. We know that Taiwan's OBM, ODM, OEM industries thrive on optimizations such as cost-cutting and battery-efficiency. To maintain this edge on a large software system like Android, we need a balanced design and thoughtful abstraction and specification: boot-up time vs. app init time, Java vs. native, fast bootup vs. apps mix, the right parameter settings, and memory management optimizations. We'll go through this kind of studies, as time allows.
講者:廖世偉 Shih-wei Liao ( Tech Lead, Google )
自介:Shih-wei Liao is a tech lead at Google Mountain View while being on-leave from National Taiwan University. He received his BS from NTU in 1991 and his MS and PhD from Stanford University in 1995 and 2000, respectively. He has many years of product experience at Intel and Google and more than two dozens of patent applications. On the other hand, he also publishes extensively on multicores, compilers, programming systems, and blocking & affine partitioning theory which is included in the Dragon Book. (Dragon book is a bible in the field of compilers.)
摘要:How to systematically achieve smaller Android footprint and optimized Android execution? I'll introduce the design of the Android toolchain and present several case studies on optimizations. We know that Taiwan's OBM, ODM, OEM industries thrive on optimizations such as cost-cutting and battery-efficiency. To maintain this edge on a large software system like Android, we need a balanced design and thoughtful abstraction and specification: boot-up time vs. app init time, Java vs. native, fast bootup vs. apps mix, the right parameter settings, and memory management optimizations. We'll go through this kind of studies, as time allows.