Person:
Vuduc, Richard

Permanent Link
https://hdl.handle.net/1853/71806
Associated Organization(s)
ORCID
ArchiveSpace Name Record

Filters

2010 - 2013 4
3 1
3 1
4
4
Reset filters

Settings

Publication Search Results

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    Qameleon: Hardware/software cooperative automated tuning for heterogeneous architectures
    (Georgia Institute of Technology, 2013-08) Kim, Hyesoon ; Vuduc, Richard
    The main goal of this project is to develop a framework that simplifies programming for heterogeneous platforms. The framework consists of (i) a runtime system to generate code that partitions and schedules work among heterogeneous processors, (ii) a general automated tuning mechanism based on machine learning and (iii) performance and power modeling techniques and profiling techniques to aid code generation.
  • No Thumbnail Available
    Item
    How much (execution) time and energy does my algorithm cost?
    (Georgia Institute of Technology, 2012-08-24) Vuduc, Richard
    When designing an algorithm or performance-tuning code, is time-efficiency (e.g., operations per second) the same as energy-efficiency (e.g., operations per Joule)? Why or why not? To answer these questions, we posit a simple strawman model of the energy to execute an algorithm. Our model is the energy-based analogue of the time-based "roofline" model of Williams, Patterson, and Waterman (Comm. ACM, 2009). What do these models imply for algorithm design? What might computer architects tell algorithm designers to help them better understand whether and how algorithm design should change in an energy-constrained computing environment?
  • Thumbnail Image
    Item
    A Roofline Model of Energy
    (Georgia Institute of Technology, 2012) Choi, Jee Whan ; Vuduc, Richard
    We describe an energy-based analogue of the time-based roofline model of Williams, Waterman, and Patterson (Comm. ACM, 2009). Our goal is to explain—in simple, analytic terms accessible to algorithm designers and performance tuners—how the time, energy, and power to execute an algorithm relate. The model considers an algorithm in terms of operations, concurrency, and memory traffic; and a machine in terms of the time and energy costs per operation or per word of communication. We confirm the basic form of the model experimentally. From this model, we suggest under what conditions we ought to expect an algorithmic time-energy trade-off, and show how algorithm properties may help inform power management.
  • Thumbnail Image
    Item
    Algorithms and software with turnable parallelism
    (Georgia Institute of Technology, 2010-09-30) Vuduc, Richard