Practice 10 questions on ELEC370 - Electrical Engineering Principles at Harvey Mudd College. Free AI-generated quiz on uNotes — track your score, retake anytime.
1According to the course materials, why is 'speculate-and-squash' not guaranteed to perform better than 'detect-and-stall' for control hazards?
2What are the primary advantages of resolving data hazards using the 'Detect and Forward' method compared to other methods mentioned?
3What is the primary cause of data hazards in a pipelined processor?
4In a 32-bit, byte-addressable architecture, what is the total size (in bytes) of a struct containing an int (4 bytes), a double (8 bytes), and a char (1 byte), assuming optimal reorganization to minimize size?
5Given a processor where 15% of instructions are branches and 75% of those branches are taken, what is the CPI contribution from control hazards if the branch misprediction penalty is 3 cycles (assuming predict-not-taken)?
6In the LEGv8 assembly code optimization example, how can a value in register X3 be multiplied by 127 using only two instructions without an explicit multiply command?
7What does the 'gap' between 32-bit IEEE floating point numbers represent?
8When applying caller-saved conventions to the variable 'e' in the foobar() loop example, what was the estimated total number of loads/stores?
9Which of the following pipeline stages is responsible for reading from the data memory in the standard 5-stage LC2K pipeline?
10If a system uses Little Endian memory organization and address 0x1000 contains 0x4B and address 0x1001 contains 0x12, what is the 16-bit hex value represented when reading from 0x1000?
