Submission Guidelines
The CoDAIM 2026 workshop seeks research papers and position articles on innovations at the intersection of AI, systems, and architecture. We invite short (up to 2-page) submissions that present early-stage ideas, empirical results, or visionary perspectives.
Accepted papers will be presented as 15-minute talks (12 minutes presentation + 3 minutes Q&A) and included in the workshop proceedings.
Paper Structure
Papers should include the following sections:
- Problem Statement: What's the problem? Include motivating evidence.
- State-of-the-Art: What are the current approaches?
- Proposed Solution: What is your solution/key insight?
- Quantitative Evidence: What evidence shows why the solution works?
- (Optional) Discussion/Future Work
- AI Use Statement: How was AI used in this paper? (not counted toward page limit)
All main content should fit within 2 pages. References are unlimited and do not count toward the page limit.
Template
Please use the ASPLOS submission template for your paper by default. Paper submissions will be single-blind.
Topics of Interest
We welcome submissions on co-design across the full stack: Application + Algorithm + PL + OS + Network + Compiler + Architecture. Topics include, but are not limited to:
Systems for AI
- Foundations of generative and agentic AI inference
- Architectural patterns and design considerations for AI systems
- Cost-, latency-, and energy-optimized inference techniques
- System architectures for serving AI agents
- Case studies from real-world deployments
- Bottleneck identification and performance tuning in model serving
- Model compression (quantization, pruning, distillation)
- Caching, batching, and speculative decoding for throughput optimization
- Hardware-aware and system-level optimizations (e.g., FlashAttention, scheduler tuning)
- Efficient inference on specialized hardware (GPUs, TPUs, custom ASICs)
- Deployment strategies for large-scale agentic systems
- Evaluation methodologies for inference optimization
- Scaling laws and trends in inference workloads
AI for Systems
- AI-driven system optimization and auto-tuning
- Machine learning for datacenter and edge scenarios
We welcome both academic and industry submissions.
Review Process
Each paper will be reviewed by the program committee, including at least two senior reviewers (e.g., faculty member, senior industry researchers). Reviews will include an accept/reject recommendation (score 1-5), strengths and weaknesses, and feedback for developing the submission into a full conference paper.
Student Mentorship
To foster mentorship, each PC member may assign up to one student to help review papers. Students must be nominated upon accepting the PC invitation. The final review must be written by the PC member. After reviews are complete, we will formally recognize student reviewers for their contributions.
Important Dates
AI Use Policy
Authors may leverage AI tools for innovative use-cases including, but not limited to:
- Polishing text and grammar
- Writing scripts to plot graphs
- Writing and debugging implementation code
However, AI tools should not be used to autonomously generate an entire paper (from coming up with an idea to writing the paper).
Each submission must include a brief statement (not counted toward the page limit) describing how AI tools were used.