Published April 20, 2026
| Version v3
Software
Open
COMET: A Cooperative Scheduling Framework for Concurrent PIM/CPU Execution on Mobile Devices
Authors/Creators
-
1.
Shanghai Jiao Tong University
-
2.
-
3.
-
4.
Description
This artifact provides the complete experimental framework for COMET from ISCA'26, a cooperative scheduling framework designed for concurrent PIM/CPU execution on mobile devices.
The artifact is built on the Ramulator2 simulator, which has been extended with the proposed low-interference PIM control interface and idle-aware scheduling methodology.
The software components include the modified C++ source code of the Ramulator2-based memory controller, specialized PIM instruction support, and the scheduling logic for PIM Execution Queues (PEQ) and PIM Read/Write Queues (PRWQ).
To represent realistic mobile scenarios, we provide two types of memory traces:
1) LLM workload traces (e.g., DeepSeek-R1-1.5B ) generated via Python scripts to simulate PIM tasks, and 2) Mobile application traces (e.g., Browser, YouTube ) collected from physical mobile devices to represent CPU tasks.
This artifact is designed to reproduce the key results presented in the paper, specifically the performance improvements and CPU interference analysis shown in Figures 8 and 10.
Users can expect to observe that COMET significantly enhances PIM throughput (up to 2.8\texttimes ) while maintaining minimal CPU performance loss (less than 5.0\% ).
The artifact requires a GCC 12+ compiler for building the simulator environment.
Files
ramulator2-sim-clean.zip
Files
(14.9 MB)
| Name | Size | Download all |
|---|---|---|
|
md5:8868a5aa27dc246941d9d9d65491f2d2
|
14.9 MB | Preview Download |