Question

1a. convert the following decimal number to 32 bit single precision Floating point binary number and convert that binary number to hexadecimal

NUMBER = -134.5 in decimal

b. convert the following 32-bit single precision floating point number to decimal:

01000111111100000000000000000000

2. Using Booth's algorithm, multiply the decimal numbers -12 and +13.

3. you have two improvement alternatives, which is better and why?

The first one improves 15% of the instructions, and it improves that speed by a factor of 6.

The second one improves 20% of the instructions, and it improves that speed by a factor of 5.

4. Using the Hennessey and Patterson FP addition method, add 1.11x2^3 to 1.01 x 2^4

5. If you have 64 sets, 8 tag bits and 4 bits for the block offset:

A. what is the set associativity?

B . what is the block size?

FIBs: CHOOSE 5

1. Calculate the Execution time [old], when the speedup overall is 2.8 and the execution time [new] is 1 ps

2. Which is better and why: a miss rate of 0.01 and a miss rate of 0.02?

3. If you have a direct mapped cache with 512 blocks, where will a main memory location of 6478 be placed?

4. Describe what might happen if the cache fills up, and you need to place something into it.

5. Calculate the average memory access time when 20% of the program instructions are memory accesses, the hit time is 2 cycles, the miss time is 16 cycles, and the miss rate is 0.02.

6. If the main memory address space is addressed with 64 bits, what is the size of it?

7. Define temporal locality and provide an example

8. What happens to the dynamic power required for a chip if you double the frequency?