🏆 Distinguished Work Award for Professor Lana Josipović and Jiantao Liu at ISCA 2025
In the world of electrical engineering and computer engineering, innovation in the design of digital systems and processors becomes a vital topic for driving technological development. Professor Lana Josipović, the leader of the DYNAMO team specializing in digital design systems and automation, and researcher Jiantao Liu received a distinguished award at ISCA 2025 (the International Symposium on Computer Architecture), in recognition of their contribution to developing in-memory processing acceleration solutions using Integer Linear Programming.
⚡ Technical summary: The award carries great importance in the field of digital processor design, as the research addressed the use of Integer Linear Programming techniques to improve the performance of Processing-in-Memory (Processing-in-Memory – PIM) acceleration. This approach addresses a traditional problem represented by the slowdown in data transfer between the processor and memory, which is a key challenge in improving efficiency and power in modern devices.
🔧 What is Processing-in-Memory?
Traditionally, the central processing unit (CPU) and memory are separated by a data transfer unit called the data bus, and thus the processor reads and writes data between two separate units. The major problem here is latency and reduced efficiency in data transfer, especially when dealing with massive amounts of data, as in artificial intelligence and signal processing.
Processing-in-Memory (PIM) is a technology based on integrating computing processing capabilities into the memory unit itself, or close to it, which reduces the need to move data to processing units and thus increases operational speed and energy efficiency.
- Reducing data transfer time between the processor and memory.
- Lowering the energy consumption caused by data movement.
- Suitable for data-intensive applications such as artificial intelligence.
- Overall increase in system performance.
🔹 Important point: PIM technology requires advanced design solutions to ensure task scheduling in a way that achieves the best possible use of available resources.
📊 Integer Linear Programming and Its Role in Improving PIM
Integer Linear Programming is based on a mathematical model through which an objective function is maximized or minimized, along with a set of conditions and variables that take integer values. In the context of Processing-in-Memory acceleration, this technique is used to plan execution schedules, distribute tasks, and select resources in a way that achieves the best performance based on the hard constraints of the processing and memory system.
- Efficiently determining in-memory processing operation schedules.
- Managing the distribution of digital resources within computing memory.
- Improving optimal utilization of integrated processing capabilities.
Applying this approach in PIM system design enhances performance without the need for expensive additional hardware.
🛡️ The Importance of the Award and Its Role in Engineering Education
The Distinguished Artifact Award is granted in recognition of projects and research that provide well-crafted and innovative tools, models, or technical solutions that achieve a tangible practical impact in their field. Awarding this prize to Professor Josipović and Jiantao Liu confirms the importance of combining computational mathematical methodologies with circuit and digital system designs to improve performance in modern devices.
For students, technicians, and trainees in the fields of electrical and digital engineering, this research offers a practical example of:
- How digital design can be connected with mathematical programming to improve computing systems.
- The role of optimal analysis in utilizing hardware resources.
- The importance of research and development in overcoming traditional performance constraints in processing systems.
📌 Quick takeaway: The award highlights the direction of electrical engineering toward integrating computing and high-performance requirements through advanced mathematical tools as part of engineering solutions.
⚠️ Important Practical and Field Applications
Among the fields that benefit from PIM technologies and solutions optimized by linear programming are:
- High-performance computing systems that depend on processing massive amounts of data.
- Developing integrated circuits for artificial intelligence and cloud computing systems.
- Smart network management systems that need faster processing and improved energy consumption.
- Designing integrated processing units used in electrical load modeling and time-based calculations.
Learning these technologies contributes to increasing the ability of engineers and technicians to innovate integrated and efficient solutions in multiple fields.
📐 The Impact on Educational and Training Curricula
The concepts of in-memory acceleration and optimal planning techniques should be included in engineering education curricula, with emphasis on:
- Providing students with analytical and mathematical tools that enable them to design complex systems.
- Developing skills in dealing with digital systems and physical devices in an integrated manner.
- Encouraging applied scientific research in graduation projects and vocational training.
⚡ Educational hint: Linear programming tools and engineering software can be used to simulate and analyze digital processor designs within an interactive educational environment.
🔍 Conclusion
The development of in-memory processing acceleration systems using Integer Linear Programming represents an important qualitative step in the field of electrical engineering and computer engineering. Winning the Distinguished Artifact Award for this scientific work shows the importance of innovation in combining mathematical programming with digital design to achieve higher performance and improved efficiency.
The most important message for students and technicians is to focus on understanding the mathematical and technical principles behind system design, and to learn how to use advanced tools that enable them to build real solutions to complex problems in the world of energy, computing, and load management.
📌 Quick takeaway: Scientific achievements such as the OptiPIM project form a successful model that connects theory with professional practice in electrical engineering disciplines.
📚 Related articles
Discover more from Mohdbali
Subscribe to get the latest posts sent to your email.
