Micro program executes the OPCODE. These programmes are available in ROS of the processor.

Micro programs are written to produce relevant digital signals of the given OPCODE. Hence, option (2) is correct.

Micro programs are written to execute the OPCODE of the instruction. Epodes are identified, and then appropriate signals for registers, accumulator, program counter, etc. are generated. These are digital signals. 

Tristate buffers have three outputs; i) 1  ii) 0  iii) open circuit (high resistance). When device is not in use, it goes in tristate condition, i.e. open circuit. This condition isolates the device from the bus. 

Micro programs slow down the processing speed due to overheads attached with accessing them compared to direct hardware execution as done in CISC. 

Micro programs allowed more than one action in a single OPCODE, thereby reducing the circuits required compared to CSIC. This helped in execution of more complex instructions in reduced circuits. So, ASIC became popular.

Cycle stealing is performed during the execution cycle of CPU. At this time, CPU does not access the memory. Hence, DMA takes over the memory access operation to move data within memory. 

DMA uses the system bus for accessing the memory cells directly. So, CPU is not involved in this operation. This is possible only when CPU is not using the system bus. In execution cycle, CPU does not use the system bus. Hence, DMA does not affect the CPU and performs its operation. 

Devices connected to I/O terminals are normally slower in access compared to the processor. So, data on the terminals may be available before the device accesses them. This data must remain on the terminal till device accesses it. This is possible only when the data is latched on the terminal. So, I/O are latched.

Flag is a flip flop inside the status register while semaphore is a memory cell within the programme definition. So, flag is hardwired and fixed while semaphore is software bit, and can be defined according to programme. Option 1 is correct.

DMA does only memory access. It does not execute instructions while co-processor is complete processor for its dedicated program. So, it executes instructions. 

Overlay concept uses the free blocks of the main memory during run time to execute some portion of a program. So, when a larger size program is required to be executed, it is first fragmented in smaller continuous programmes, and those programmes are loaded and removed in/from memory as and when required.

Networking is mainly provided by the LAN-card, OS and other software. So, a general purpose processor can handle the networking.

In multitasking, the OS uses fixed clock for every segment of execution and executes them in cyclic manner. While in multiprogramming, another program will not be executed till CPU goes in idle cycle. Option 1 is correct.

When hands are working, eyes may look somewhere and provide visuals while feet are doing the walking. All the above operations are performed simultaneously without affecting each other and without pausing. This type of working is distributed processing. 

Cache memory is used when pipelining is implemented in the processor architecture. 80486 architecture supports the pipelining. 

ECL can be easily implemented to have high speed memory chip. Because of its speed, it is favourable for implementing cache memory. 

For parallel I/O operation, microprocessor uses a programmable peripheral interface (PPI) IC as controller/interface device for parallel I/O while this is internally implemented in microcontroller. So, microcontroller does not require 8255. 

Memory chip accepts the data bit at the input and loads it in by WR, and outputs data bit at output by RD. So, both output and input remain connected to the bus where tristate nature is required. So, these buffers are required at both input and output. 

Due to its high density, low power requirement and generally acceptable speed, MOS memories are most popularly used memories. 

Loader uses the compiled and linked program to move it from secondary storage to main memory for its execution by CPU. Hence, it must be required when the program is ready for execution. Option 3 is correct.

I/O controllers have separate processors and dedicated instruction formats to define various dedicated operations for the specific device. They offload CPU during I/O operations.

The local symbols/ functions are resolved by the compiler because all references are available to compiler. But global symbols/functions cannot be resolved by the compiler because the references are not available to the compiler. These are resolved after the compilation using linker. 

Writing a compiler requires modular and structured functions to execute compiler operations. One may also require to access the hardware directly, so such facility is required in the language. The language must also provide facility to handle complex data bases and structures with ease. C language provides all such facilities. So, it will be picked.

CPU is much faster than the I/O device. A buffer block of 4 K byte means the block can store 4K characters. Writing 4K characters in the block requires very less time for CPU while for I/O requires quite high time for accessing these 4K. CPU becomes free for sufficient time to do other part of programmes after writing a buffer, while I/O controller takes over to send characters and other commends to I/O device independent of CPU. So, the CPU time is saved.