Elements of Computing Systems Chapter 5

Joel Chippindale edited this page Feb 5, 2015 · 13 revisions
Clone this wiki locally

The gang

Preamble

Welcome Drew & Vaidas.

Vaidas managed to buy Understanding Computation :trollface: A great purchase nonetheless.

Chapter 4 and Von Neumann architectures throws the reader in at the deep end. We agreed that seeing the assembler and instruction set first clarifies the design in Chapter 5.

Is it a Von Neumann architecture or a Harvard architecture? In the book’s architecture the program data is stored in ROM and data is stored in RAM. Whereas it seems in Von Neumann the data and program reside in the same memory.

Murray reminded us that the computers that some of us started writing programs, e.g. the BBC Model B, on had their OS's on ROMs and so an OS upgrade meant replacing a chip.

Page 97/98 explains the difference between a general purpose and dedicated computer. They choose to use ROM and RAM rather than just RAM in order to have a single cycle fetch/execute logic.

Looking ahead: will Tetris be implemented in ROM. Or will our OS load it from somewhere else?

Memory mapping: what preceeded it?

What do we have to do in this chapter? To what extent will we have to think about fetch/execute?

decode is a roundrect. What does this indicate?

Exercises

Drew gets a crash course in HDL.

We took a look at our implementation of RAM16K for reference.

Another question of which is the most significant bit.

Had a brief conversation about tick/tock and whether they’re seen at all from outside.

We caused the test to hang by not implementing Keyboard in our Memory.

Turns out we’ve been getting the indexing the wrong way around all this time.

The Memory test has to be performed in the simulator because it requires us to interact (pressing keyboard and visually verifying screen).

The whiteboard

What next?

We got Memory working but ran out of time and spent a little while deciding how to proceed. The group preferred meeting again sooner to tackle the rest of the exercise with it fresh in our heads.

With this in mind we plan to have a bonus Computation Club meeting on Tuesday 10th Feb at Geckoboard for those who wish to work through the rest of the exercises from chapter 5 together.

Then normal service will resume on Tuesday 24th Feb when we will look at Chapter 6.

Thanks

Thanks to Leo for hosting us at Geckoboard's offices and thanks to Geckoboard, Jamie and Drew for providing refreshments and snacks.