3rd Workshop on Unary Computing (WUC)
January 25, 2026
February 1, 2026
Our brain operates at 20 watts, while one rack of CPUs and GPUs for similar tasks demand staggering 10,000 watts or even more. The prevalence of Generative AI is just making it worse, due to billions of model parameters. Leveraging a compute fabric based on event-driven spikes, our brain effortlessly executes complex tasks with unmatched energy efficiency, surpassing the capabilities of any existing von-Neumann architecture. These spikes manifest as unary data, conveyed as distinct 0s and 1s sequentially, diverging from the conventional binary representation. This unary computing model is omnipresent in our neurons, supporting human intelligence with extreme efficiency, with the earliest form of dates back to 1960s. Unary computing has been extensively utilized for both arithmetic and neural operations, by different groups targeting various applications, showcasing its potential in the signal processing, error correction, and AI. This workshop aims to establish connections among global computing researchers, fostering the exchange of ideas in unconventional computing paradigms, broadly centered around computing on unary data. As AI is really devouring our energy, it is crucial to contemplate the broad implications of unary computing on efficient perception, cognition, AI, and beyond. The workshop serves as a platform to explore the profound impact of unary computing in these domains.
Papers that contribute to the theory, applications and system of unary computing are solicited. General topics of interest include (not limited to):
-
Theory: new computing capability with unary encoding (rate coding, temporal coding, their mixture, etc.), unary neural processing, metric formulation, simulation framework, etc.
-
Application: brain-related cognitive tasks, biological perception tasks for embodied AI, non-bio-specific applications that can take advantage of unary encoding or unary computing, etc.
-
System: in/near-sensor unary computing circuit, unary instruction set architecture, programming language, scaling-up unary computing in systems, synergistic computing with other paradigms, etc.
Scope: Works that are unpublished, working-in-progress and published are all welcome.
Format:
-
Submission must be in English and up to 4 pages in PDF format (excluding references). Please use the sigplan proceedings template of ACM’s acmart Latex class on the official ACM site. https://www.acm.org/publications/proceedings-template
-
Submissions shall not include any author identifying information.
-
Accepted papers will be presented at the workshop and included in the workshop report.
- Submission site: https://wuc2026.hotcrp.com/
Important Dates:
-
Submission deadline: Feb 1, 2026 (AOE)
-
Acceptance date: Feb 13, 2026
-
Camera ready: March 15, 2026
-
Workshop: March 22, 2026