No abstract

No abstract

Over the past few years, there has been vast growth in the area of the web browser as an applications platform. One example of this trend is Google's Native Client (NaCl) platform, which is a software-fault isolation mechanism that allows the running of native x86 or ARM code on the browser. One of the security mechanisms employed by NaCl is that all branches must jump to the start of a valid instruction. In order to achieve this criteria though, all return instructions are replaced by a specific branch instruction sequence, which we call NaCl returns, that are guaranteed to return to a valid instruction. However, these NaCl returns lose the advantage of the highly accurate return-address stack (RAS) in exchange for the less accurate indirect branch predictor. In this paper, we propose a NaCl-RAS mechanism that can identify and accurately predict 76.9 on average compared to the 39.5 of a traditional BTB predictor.

Specialized accelerators have become prevalent in mobile computing platforms for their ability to perform certain tasks, such as image processing, at a lower power cost than CPUs or GPUs. In this paper, we focus on using Cellular Neural Networks (CNN) as a specialized accelerator. CNN is a neural computing paradigm that is well suited for image processing applications. However, hardware implementations were originally developed only to handle relatively small image sizes. In this paper, we propose SP-CNN, an architecture as well as a multiplexing algorithm that provides scalability to CNN applications. We demonstrate the proposed multiplexing algorithms over a set of six image processing benchmarks, as well as present a performance and power analysis of SP-CNN.