ECS154A-L18

📗 -> 02/19/25: ECS154A-L18

---

[Lecture Slide Link]

🎤 Vocab

❗ Unit and Larger Context

✒️ -> Scratch Notes

Performance

Can be defined by throughput, other times by response times

• Throughput: The amount of scalibility / things that can be finished on mass
  • ATMs: Customers don’t really care how much their individual time takes, but want to get as many customers processed as possible.
• Response Times: The speed you can finish something.
  • Weather forecast: A real time system where you need to get a forecast to a customer in time. Not interested in yesterday’s forecast.

Measure Performance?

• Elapsed time
• CPU time
  Ideally, we want to get a machine that’s fastest on the programs we do all the time. This is hard to measure, so we measure performance on representative programs.
  Run representative Programs
• Benchmarks: Representative of what you do with a computer

4 Types:

1. Synthetic (Dhrystrone, Whetstone, etc.)
2. Toy (Towers of Hanoi, Quicksort, etc.)
   1. Lots of recursion
3. Kernel (Livermore Loops, Lispack Loops, etc.)
4. “Real” (SPEC suite, Perfect Club, etc.)

Kernel - The part of the OS that has to remain in memory at all time. The core/essence of the OS.

Reporting Performance

MIPS
MFlop

• Mega Flop
  Elapsed Time (weights)

Improving Performance

1. Faster clock
2. Fewer clocks/inst
3. Use fewer insts

1. - compiler
2. - organization
3. - technology

🧪 -> Refresh the Info

Did you generally find the overall content understandable or compelling or relevant or not, and why, or which aspects of the reading were most novel or challenging for you and which aspects were most familiar or straightforward?)

Did a specific aspect of the reading raise questions for you or relate to other ideas and findings you’ve encountered, or are there other related issues you wish had been covered?)

🔗 -> Links

Resources

• Put useful links here

Connections

• Link all related words