ECEn 424 Review Topics: Chapter 6

---

• Memory:
  • What are the important tradeoffs between static RAM (SRAM) and dynamic RAM (DRAM)?
  • How are conventional DRAMs organized and accessed?
  • What kind of enhancements have been employed to increase the speed of DRAMs?
  • How do the various types of nonvolatile memories differ from each other?
  • What does the typical bus structure look like that connects the CPU to main memory?
• Disks:
  • How is information organized on a disk? (What are tracks, sectors, and cylinders?)
  • What is multiple zone recording?
  • How many read/write heads and actuator arms does a disk have?
  • What are the three main components of the access time for a disk sector, and what are their relative magnitudes in practice?
  • What are logical blocks?
  • When a processor wants to read from or write to a disk, what sequence of actions typically takes place?
  • What is a solid state disk, and what advantages and disadvantages does it have relative to magnetic disks?
• Over a 30-year window, what important trends can be observed regarding changes in CPU clock rates and SRAM, DRAM, and disk access times and sizes? What are the implications for system designers?
• Locality:
  • What is locality and why is it present in typical programs?
  • What distinct forms of locality exist?
• Why do modern computer systems use memory hierarchies, and what are the general principles of their design and operation?
• Caches:
  • Why are caches an important part of the memory hierarchy?
  • What determines where a block can be placed in a cache?
  • What determines which block will be chosen as the victim block for replacement?
  • What is the difference between compulsory, conflict, and capacity misses?
  • What are the key parameters that determine the organization of a cache?
  • When a memory request arrives at a cache, what steps are taken to determine if the requested block is present?
  • What are the differences between direct-mapped, set-associative, and fully-associative caches? (Which is most likely to avoid thrashing? Which is fastest? Which is the most expensive?)
  • What special challenges do writes create in cache-based systems?
  • How does one evaluate the performance of a cache?
• What kinds of changes can be made to code to make it more cache friendly? (Can you tell by inspection which version of code results in the best cache performance?)
• Given the source code to a C function that accesses a matrix and details about block size, cache size, and cache organization, can you determine how many misses will occur on the array data?
• What information can be extracted from a memory mountain for a particular machine, and how is the data for the mountain obtained?
• Self test: Can you complete all practice problems in Chapter 6?

Updated for 3rd Edition of CS:APP