The views and opinions expressed in this page are strictly those of the page author.
The contents of this page have not been reviewed or approved by the University of Minnesota.

Last updated January, MMXIX by Ulya Karpuzcu.


ALTAI lab explores how to improve the energy-efficiency of computing via

  1. pushing traditional (CMOS-based) computing to its limits

  2. computing with post-CMOS devices and paradigms


  1. July MMXVII: Karpuzcu, Sapatnekar, Wang received an NSF SPX grant on scalable in-memory processing using spintronics.

  2. January MMXVI: Karpuzcu received the NSF CAREER Award on “Trading Communication and Storage for Computation to Enhance Energy Efficiency”.

  3. June MMXV: We released ACCURAX, a compilation of safe accuracy metrics for approximate computing, along with an output randomization tool.

  4. September MMXIV: Karpuzcu received an NSF CCF grant on soft (approximate) near-threshold voltage computing.

  5. August MMXIV: Lilja, Sartori, and Karpuzcu received an NSF XPS grant on approximate communication to enhance energy efficiency.

  6. January MMXIV: The second Workshop on Near-threshold Voltage Computing will be held w/ ISCA. Call for Papers

  7. Spring MMXIV: New Course Announcement: EE8950 Physics of Computation Syllabus

  8. Inspired by Richard Feynman’s lectures, “Physics of Computation” explores how physical principles/limits shape paradigms of computing.

  9. Goal: To understand how (and to what extent) a paradigm shift in computing can help with emerging energy problems.   

  10. Scope: Physical limits of computing, coding and information theoretical foundations, computing with beyond-CMOS devices, reversible computing,

  11. quantum computing, stochastic computing. For each computing paradigm, we cover (i) how information is represented, processed, stored, and communicated;

  12. (ii) to what extent shortcomings can be addressed; (iii) how the application domain looks like.

  13. September MMXII: We are organizing the first Workshop on Near-threshold Voltage Computing w/ MICRO. Call for Papers

Recent Publications    Full List

  1. POWERT Channels: A Novel Class of Covert Communication Exploiting Power Management Vulnerabilities, HPCA 2019.

    BibTeX   CAL Version

  1. Trading Computation for Communication: A Taxonomy of Data Recomputation Techniques, IEEE TETC 2018.


  1. Toward Dynamic Precision Scaling, IEEE Micro Magazine Special Issue on Approximate Computing, 2018.


  1. StochMem: Towards Seamless Stochastic Computing Systems with Analog Memories, CAL 2018.


  1. Efficient In-Memory Processing Using Spintronics, CAL 2017.


  1. ThermoGater: Thermally-Aware On-Chip Voltage Regulation, ISCA 2017.

    Slides   BibTeX

  1. AMNESIAC: Trading Computation for Communication for Energy Efficiency, ASPLOS 2017.

    Slides   BibTeX


We released:

  1. QuantumOps, an R package for quantum computing.

  2. VARIUS(TC,NTV), a (micro)architectural model of parametric variation.

  3. ACCURAX, a compilation of safe accuracy metrics for approximate computing, along with an output randomization tool.

Information for Prospective Students

  1. Our lab’s doors are wide open to motivated Ph.D. students and undergraduate researchers.

  2. For M.S. students, the thesis option is a necessary but not sufficient condition.


Our research has been funded by

Main            Publications            People            Open-Source            Courses Offered       


University of Minnesota, Twin Cities, Department of Electrical and Computer Engineering