Synthesis of Bayesian Machines On FPGAs Using Stochastic Arithmetic
نویسندگان
چکیده
Probabilistic inference allows artificial systems to cope with uncertainty, but it can be computationally demanding. Inspired by biological neural systems, stochastic arithmetic modules on reconfigurable hardware can provide massively parallel systems with limited resources. This work presents a framework to automatically implement Bayesian Machines to perform computations using stochastic bitstreams.
منابع مشابه
Dynamical Control of Computations Using the Family of Optimal Two-point Methods to Solve Nonlinear Equations
One of the considerable discussions for solving the nonlinear equations is to find the optimal iteration, and to use a proper termination criterion which is able to obtain a high accuracy for the numerical solution. In this paper, for a certain class of the family of optimal two-point methods, we propose a new scheme based on the stochastic arithmetic to find the optimal number of iterations in...
متن کاملImproved Synthesis of Compressor Trees on FPGAs by a Hybrid and Systematic Design Approach
Improving arithmetic circuits on FPGAs is one of the main imperatives of FPGA vendors. Augmenting logic cells with dedicated arithmetic components such as adders and carry chains indicates the need for such improvements. In a prior work, we showed how the carry chains in the state-of-the-art Altera FPGAs could be exploited for synthesis of compressor trees. In that work, we proposed generalized...
متن کاملFPGA-specific synthesis of loop-nests with pipelined computational cores
The increased capacity and enhanced features of modern FPGAs opens new opportunities for their use as application accelerators. However, for FPGAs to be accepted as mainstream acceleration solutions, long design cycles must be shortened by using high-level synthesis tools in the design process. Current HLS tools targeting FPGAs have several limitations including the inefficient use of deeply pi...
متن کاملClosing the Gap between FPGA and ASIC: Balancing Flexibility and Efficiency
Despite many advantages of Field-Programmable Gate Arrays (FPGAs), they fail to take over the IC design market from Application-Specific Integrated Circuits (ASICs) for high-volume and even medium-volume applications, as FPGAs come with significant cost in area, delay, and power consumption. There are two main reasons that FPGAs have huge efficiency gap with ASICs: (1) FPGAs are extremely flexi...
متن کاملFPGA Based Custom Computing Machines for Irregular Problems
Over the past few years there has been increased interest in building custom computing machines (CCMs) as a way of achieving very high performance on specific problems. The advent of high density field programmable gate arrays (FPGAs), in combination with new synthesis tools, have made it relatively easy to produce programmable custom machines without building specific hardware. In many cases, ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2014