Digital Design & Comp. Arch: L29: Problem Solving III (Spring 2025)

Digital Design & Comp. Arch: L29: Problem Solving III (Spring 2025)

by Onur Mutlu

Formal & Physical Sciences Computing & Cybersecurity UYComputer scienceUYFComputer architecture and logic design
🎙 Onur Mutlu 👥 64K 📅 July 17, 2026 ⏱ 171 min 👁 484 📄 tutorial 🧭 2026-08-15
Available in: English (current) Français

Keywords

branch predictionsystolic arrayGPUSIMDcache

Summary

This lecture, part of the Digital Design and Computer Architecture course at ETH Zurich, focuses on solving exam-style problems from previous years. The instructor, Prof. Onur Mutlu, guides students through a series of questions covering branch prediction, systolic arrays, GPU and SIMD processing, vector processing, cache tracing, memory hierarchy, prefetching, and cache performance analysis. The session begins with a problem on locally and globally correlated branches, explaining how to identify them and calculate the behavior of a global branch predictor. Next, a systolic array design for matrix multiplication is discussed, detailing the operations of processing elements and the arrangement of inputs. The lecture then addresses GPU SIMD utilization, considering the impact of divergent branches on efficiency. Vector processing is explored, focusing on memory bank conflicts and stall avoidance. Later problems involve tracing cache behavior, analyzing memory hierarchy performance, and understanding prefetching techniques. The session concludes with advanced topics like reverse engineering caches. Throughout, the instructor emphasizes problem-solving strategies and the underlying principles of computer architecture.

165 words

Critical Evaluation

Value of the Information & Strength of the Argument

The lecture provides high-value information by walking through complex problems step-by-step, clarifying common misconceptions, and reinforcing fundamental concepts. The argumentation is solid, as each solution is logically derived from first principles and the given specifications. The instructor’s explanations are thorough, making the reasoning transparent and educational. The problems are well-chosen to illustrate key aspects of branch prediction, systolic arrays, GPU execution, and memory systems, offering practical insights into real-world architectural design.

Scientific Rigor, Source Quality, Title Accuracy

The scientific rigor is high, as the lecture is based on a university course and references academic papers and course materials. The sources cited in the description are relevant and credible, including papers on processing-in-memory and RowHammer. The title accurately reflects the content, which is a problem-solving session. The lecture is well-structured, with clear explanations and a focus on technical accuracy. The use of exam questions from previous years adds authenticity and demonstrates the application of theoretical knowledge.

164 words

Title / Content Match

The title accurately reflects the content: a lecture focused on solving problems in digital design and computer architecture.

Quality & Reliability

9/10

Lecture by a renowned professor in computer architecture, based on a structured course at ETH Zurich. The content is technical, detailed, and includes problem-solving exercises with clear reasoning. The provided sources are academic papers and course materials, indicating high reliability.

Chapters

Cited Sources

Concurring Sources

External References

Contribution & Novelties

This lecture provides a comprehensive walkthrough of challenging problems in computer architecture, offering insights into problem-solving strategies and common pitfalls. It bridges theoretical concepts with practical applications, such as designing systolic arrays and optimizing GPU utilization. The session is particularly valuable for students preparing for exams, as it demonstrates how to approach complex questions systematically.

Pour aller plus loin :

  • Branch predictor — Overview of branch prediction techniques.
  • Systolic array — Explanation of systolic array architectures.
  • SIMD — Introduction to SIMD processing.
  • GPU — General information on GPUs.
  • Cache — Fundamentals of caching.

93 words

Radar Profile

The radar profile shows high scores across all dimensions, indicating a well-rounded and authoritative lecture. The strong technical depth and reliability make it an excellent resource for advanced students.

Reliability 9/10