ComputingComputer EngineeringComputer ScienceHardware-software Co-designHeterogeneous ComputingParallel ComputingData ParallelismDirect Memory AccessHardware AccelerationKernel DesignMemory Hierarchy OptimizationMemory Subsystem DesignCode Size OptimizationCommunication-computation OverlapData AccessDatapath DesignDatapath GenerationKernel OptimizationKernel Performance OptimizationRegister File Design
This paper presents ePUMA, a master-slave heterogeneous DSP processor for communications and multimedia. We introduce the ePUMA VPE, a vector processing slave-core designed for heavy DSP workloads and demonstrate how its features can used to implement DSP kernels that efficiently overlap computing, data access and control to achieve maximum datapath utilization. The efficiency is evaluated by implementing a basic set of kernels commonly used in SDR. The experiments show that all kernels asymptotically reach above 90% effective datapath utilization. while many approach 100%, thus the design effectively overlaps computing, data access and control. Compared to popular VLIW solutions, the need for a large register file with many ports is eliminated, thus saving power and chip area. When compared to a commercial VLIW solution, our solution also achieves code size reductions of up to 30 times and a significantly simplified kernel implementation.
4
Shekhar Borkar, Andrew A. Chien · Communications of the ACM · 2011 · 958 citations
Understanding sources of inefficiency in general-purpose chips
Rehan Hameed, Wajahat Qadeer, Megan Wachs et al. · 2010 · 394 citations
From SODA to scotch: The evolution of a wireless baseband processor
Mark Woh, Yuan Lin, Sangwon Seo et al. · 2008 · 101 citations
Mark Woh, Sangwon Seo, Scott Mahlke et al. · 2009 · 81 citations