Gradient descent-based programming of analog in-memory computing cores
Abstract
Gradient descent-based programming of analog in-memory computing coresJulian Büchel a, Athanasios Vasilopoulos a, Benedikt Kersting a, Frederic Odermatt a, Kevin Brew b, Injo Ok b, Sam Choi b, Iqbal Saraf b, Victor Chan b, Timothy Philip b, Nicole Saulnier b, Vijay Narayanan c, Manuel Le-Gallo Bourdeau a, Abu Sebastian aa IBM Research Europe – Zurich, Säumerstrasse 4, Ruschlikon, Switzerlandb IBM Research Albany, NY, USAc IBM Research Yorktown Heights, NY, USAProceedings of Neuromorphic Materials, Devices, Circuits and Systems (NeuMatDeCaS)València, Spain, 2023 January 23rd - 25thOrganizers: Rohit Abraham John, Irem Boybat, Jason Eshraghian and Simone FabianoContributed talk, Julian Büchel, presentation 031DOI: https://doi.org/10.29363/nanoge.neumatdecas.2023.031Publication date: 9th January 2023The precise programming of crossbar arrays of unit cells is crucial for obtaining high matrix-vector-multiplication (MVM) accuracy in analog in-memory computing (AIMC) cores. We propose a radically different approach based on directly minimizing the MVM error using gradient descent with synthetic random input data. Our method significantly reduces the MVM error compared with conventional unit cell by unit cell iterative programming. It also eliminates the need for high-resolution analog-to-digital converters (ADCs) to read the small unit-cell conductance during programming. Our method improves the experimental inference accuracy of ResNet-9 implemented on two phase-change memory (PCM)-based AIMC cores by 1.26%. © FUNDACIO DE LA COMUNITAT VALENCIANA SCITOnanoGe is a prestigious brand of successful science conferences that are developed along the year in different areas of the world since 2009. Our worldwide conferences cover cutting-edge materials topics like perovskite solar cells, photovoltaics, optoelectronics, solar fuel conversion, surface science, catalysis and two-dimensional materials, among many others.nanoGe Fall MeetingnanoGe Fall Meeting (NFM) is a multiple symposia conference celebrated yearly and focused on a broad set of topics of advanced materials preparation, their fundamental properties, and their applications, in fields such as renewable energy, photovoltaics, lighting, semiconductor quantum dots, 2-D materials synthesis, charge carriers dynamics, microscopy and spectroscopy semiconductors fundamentals, etc.nanoGe Spring MeetingThis conference is a unique series of symposia focused on advanced materials preparation and fundamental properties and their applications, in fields such as renewable energy (photovoltaics, batteries), lighting, semiconductor quantum dots, 2-D materials synthesis and semiconductors fundamentals, bioimaging, etc.International Conference on Hybrid and Organic PhotovoltaicsInternational Conference on Hybrid and Organic Photovoltaics (HOPV) is celebrated yearly in May. The main topics are the development, function and modeling of materials and devices for hybrid and organic solar cells. The field is now dominated by perovskite solar cells but also other hybrid technologies, as organic solar cells, quantum dot solar cells, and dye-sensitized solar cells and their integration into devices for photoelectrochemical solar fuel production.Asia-Pacific International Conference on Perovskite, Organic Photovoltaics and OptoelectronicsThe main topics of the Asia-Pacific International Conference on Perovskite, Organic Photovoltaics and Optoelectronics (IPEROP) are discussed every year in Asia-Pacific for gathering the recent advances in the fields of material preparation, modeling and fabrication of perovskite and hybrid and organic materials. Photovoltaic devices are analyzed from fundamental physics and materials properties to a broad set of applications. The conference also covers the developments of perovskite optoelectronics, including light-emitting diodes, lasers, optical devices, nanophotonics, nonlinear optical properties, colloidal nanostructures, photophysics and light-matter coupling.International Conference on Perovskite Thin Film Photovoltaics Perovskite Photonics and OptoelectronicsThe International Conference on Perovskite Thin Film Photovoltaics Perovskite Photonics and Optoelectronics (NIPHO) is the best place to hear the latest developments in perovskite solar cells as well as on recent advances in the fields of perovskite light-emitting diodes, lasers, optical devices, nanophotonics, nonlinear optical properties, colloidal nanostructures, photophysics and light-matter coupling.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Related challenges for in-memory computing are achieving higher computational precision, inherent stochastic switching of devices and their impacts on reliability, and levels of parallelism.
“Related challenges are achieving higher computational precision [50, 51], inherent stochastic switching of devices and their impacts on reliability [52, 53], and levels of parallelism [54].”
From Brain-inspired computing systems: a systematic literature review (2024) · cites this paper as [51]
Cited by
Sorted by citations
Brain-inspired computing systems: a systematic literature review
Zolfagharinejad, M., Alegre-Ibarra, U., Chen, T., Kinge, S., & van der Wiel, W. G. (2024) · The European Physical Journal B · doi:10.1140/epjb/s10051-024-00703-6 · Cited by 39 · see citing works
- Authors
- Büchel, J., Sebastian, A., Kersting, B., Odermatt, F., Brew, K., Ok, I., Choi, S., Vasilopoulos, A., Chan, V., Philip, T., Saulnier, N., Narayanan, V., Le-Gallo Bourdeau, M., & Saraf, I.
- Affiliations
- IBM Research - Zurich (CH)
- IBM Research Europe – Zurich
- IBM Research - Albany (US)
- IBM (United States) (US)
- ORCID
- Venue
- Proceedings of the Neuromorphic Materials, Devices, Circuits and Systems
- Published
- 2023-01-09
- OpenAlex
- W4318817284
- Cited by
- 1
- References
- 0
- Citations by year
- 2024: 1
- FWCI
- 0.49
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- Neuromorphic Materials, Devices, Circuits and Systems (NeuMatDeCaS) (publishedVersion)
- Topics
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Advanced Memory and Neural Computing
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Ferroelectric and Negative Capacitance Devices
- Physical Sciences › Engineering › Biomedical Engineering › Analog and Mixed-Signal Circuit Design
- Keywords
- analog in-memory computing, crossbar arrays, matrix-vector multiplication, phase-change memory, analog-to-digital converter elimination, neuromorphic hardware, neuromorphic computing
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review