GATE CSE 2018 | Question: 5

Question

Consider the following processor design characteristics:

1. Register-to-register arithmetic operations only
2. Fixed-length instruction format
3. Hardwired control unit

Which of the characteristics above  are used in the design of a RISC processor?

• A. I and II only
• B. II and III only
• C. I and III only
• D. I, II and III

Comments

1. Register-to-register arithmetic operations only

Some people are thinking this statement is wrong. But notice keyword arithmetic.

---

register to register can happen for logic/shift operations also, right ?

---

@Shaik Masthan

Sir, except load/store all the other operations in RISC will be  reg-to-reg only right?

Answer 1

(D) All of these 

Hardwired control units are implemented through use of combinational logic units, featuring a finite number of gates that can generate specific results based on the instructions that were used to invoke those responses. Their design uses a fixed architecture—it requires changes in the wiring if the instruction set is modified or changed. This architecture is preferred in reduced instruction set computers (RISC) as they use a simpler instruction set.

Instructions length cannot vary in RISC usually it's $32$ bit. For CISC it can be between $16$ to $64$ bits. 

The hardwired control unit is used when instructions are fixed. 

Register to register operations is always possible in RISC. CISC can have memory to memory instructions also. 

References:

•  https://www-cs-faculty.stanford.edu/~eroberts/courses/soco/projects/2000-01/risc/risccisc/
• https://en.wikipedia.org/wiki/Control_unit#Hardwired_control_unit

Comments

They have used "only" keyword in first. RISC can have other kind of instructions as well. I think it should be B.

---

According to the link you shared, the emphasis of RISC is on software than on hardware. How is that justifying the "Hardwired control unit" point?

---

@Akhilesh
 Some RISC instructions allow you to specify a signed constant, or “immediate” value, for the second source instead of a register.
RISC is still considered a load/store architecture, because arithmetic operands cannot be from arbitrary memory locations. They must either be registers or constants that are embedded in the instruction.
See this: https://courses.cs.washington.edu/courses/cse378/07au/lectures/L1.pdf (Page 9)

---

@habedo007
Hardwired control units are implemented through use of combinational logic units, featuring a finite number of gates that can generate specific results based on the instructions that were used to invoke those responses. Hardwired control units are generally faster than microprogrammed designs. Their design uses a fixed architecture—it requires changes in the wiring if the instruction set is modified or changed. This architecture is preferred in reduced instruction set computers (RISC) as they use a simpler instruction set
Source: https://en.wikipedia.org/wiki/Control_unit#Hardwired_control_unit

---

@Akhilesh Singla

This question is directly taken from Morris Mano book.

Answer 2

Answer is D

I collected information from multiple sources and created this table. If you find anything wrong, please tell in the comment section.

CISC Vs RISC

Sl.no: Parameter/Feature RISC CISC 1 Abbreviation Reduced Instruction Set Computer Complex Instruction Set Computer 2 Clock Cycle Per Instruction (CPI) 1(except LOAD and STORE) Multiple 3 Average Clock Cycle Per Instruction (CPI) 1.5 2 to 15 4 Instruction Format Fixed(usually 32 bits) Varies 5 Performance Optimization Software-centric approach Hardware-centric approach 6 Coding More lines of Code Simple 7 Code Expansion Can be Problematic Not a problem 8 Complexity Lies in compiler Lies in microprogram 9 Instruction Set Has Primitive instructions Has variety of different instructions 10 Complex Addressing Modes Synthesized using the software Already supports complex addressing modes 11 Register Sets Multiple Unique 12 Total Registers More than 256 Typically 6 to 16 13 Register Operands Per Instructon Upto 3 Upto 2 14 No Of Instructions Less More 15 Instructions Accesing Memory Only LOAD and STORE Almost all from same set 16 RAM Usage Heavy use of RAM(can cause bottleneck if RAM is limited) Comparatively more efficient 17 Hardware Design Focus Hardwired control Micro-programmed control 18 Pipelining Very High Rare 19 Instruction Decoding Simple Complex 20 Transistors Usage For memory registers For storing complex instructions 21 Computation Operands Register and Immediate only Memory operands allowed 22 Disk Space Space is saved Space is wasted 23 Stalling Mostly reduced in processors Processors often stall 24 Portability Less portable for design errors Significant portable for design errors 25 Calculation Speed Fast Slow 26 Design Time Less Long 27 Memory Unit Absent(uses a separate hardware to implement instructions) Present 28 External Memory For Calculation Purpose Not required Needed 29 General Purpose Registers 32 to 192 with split data cache for instruction cache 8 to 24 general purpose registers with a unified cache for instructions and data (recent designs use split caches) 30 Memory Size Of Program More space Less space 31 Interruption Responds to interrupt only at the proper place in instruction execution Responds to an interrupt only at the end of execution 32 Conditional Jump Based on a bit anywhere in memory Based on status register bit 33 Instruction Format Field Placement Varies Regular Consistent placement of fields 34   Register to register: "LOAD" and "STORE" incorporated in instructions Memory-to-memory: "LOAD" and "STORE" are independent instructions 35 Eg: A = A * B; <<======this is C statement LOAD R1, A          <<<======this is assembly statement LOAD R2, B          <<<======this is assembly statement PROD A, B          <<<======this is assembly statement STORE R3, A          <<<======this is assembly statement MULT A,B <<<======this is assembly statement 36 Processors Ideal for processors performing dedicated operations Ideal for processors performing a variety of operations 37 Instruction Fetching Time Mostly same time Different 38 Operands Per Instruction Fixed Variable 39 Applications High-end applications such as video processing, telecommunications and image processing Low-end applications such as security systems, home automation, etc. 40 Example Alpha, ARC, ARM, AVR, MIPS, PA-RISC, PIC, Power Architecture, and SPARC System/360, VAX, PDP-11, Motorola 68000 family, AMD and Intel x86 CPUs

 

Comments

Hope you know not every source is reliable.

---

Whats the feature or parameter name for point 34 @Balaji Jegan

---

Please elaborate point 5.

Answer 3

RISC is the traditional load/store architecture. It needs all the operands to be loaded in the registers first, and then the computations can be performed on them. (In CISC you don't need to load an operand in a register, you can directly access it from the memory)

So, I is correct.

 

RISC uses fixed-length instructions, while CISC uses variable-length instructions. For this reason, decoding an instruction is easier in RISC than in CISC.

So, II is correct.

 

RISC is more software-centric, while CISC is more Hardware-centric, BUT that doesn't mean it is dichotomous. Both architectures still need both software and hardware facilities.
For hardware, RISC has a hardwired control unit; whereas CISC has a micro-programmed control unit.
Hardwired CU is one of the consequences of the fixed-length instructions.

So, III is correct

 

Option D

Comments

Among all the answers present, this one is the best and easy to understand…

---

Best, as always :)

Answer 4

All of these

Instructions length cannot vary in RISC usually it's 32 bit. For CISC it can be between 16 to 64 bits.

Hardwired control unit are useful when instructions are fixed.

Register to register operations are always possible in RISC. CISC can have memory to memory instructions also.

Ref: https://www-cs-faculty.stanford.edu/~eroberts/courses/soco/projects/2000-01/risc/risccisc/

Comments

Register to register operations are always possible in RISC but in question it is mentioned only right ??

are you sure ?