17.
Typical Microprogrammed unit
Advantages of microprogrammed unit
18.
Micro Instructions
Horizontal micro instructions
Vertical micro instructions
Vertical micro instructions Vs Horizontal micro instructions
20.
The Memory Hierarchy/ Hierarchy memory ( use both terms atleat 3 times )
Registers
Primary Memory
Secondary Memory
Portable Secondary memory
Cache Memory
Memory Characterstics
Location
Storage capacity
Transfer modes
Access Modes
Physical properties
And few more properties – refer pdf page-4,5 of 20th video
TYPICAL MICROPROGRAMMED CONTROL UNIT
1) A microprogrammed control unit uses microinstructions in software to produce control signals.
2) A set of instructions makes up a programme.
3) A series of Micro-Operations are necessary for an instruction.
4) Control signals carry out Micro-Operations.
5) Micro-instructions are used in this instance to produce the control signals.
This implies that each instruction must be followed by a sequence of micro-instructions.
This is referred to as its micro-program.
8) In a tiny memory called the “Control Memory,” all instructions are microprogrammed.
9) The processor contains the Control memory.
10) Take into account an Instruction that is loaded into the Instruction Register from the main memory (IR).
11) The processor utilises its distinctive “opcode” to determine the address of the initial microinstruction.
12) The Control Memory Address Register, often known as pIR, receives that address.
13) To locate the relevant p-instruction from the Control Memory, this address is decoded.
14) The size of the micro-instructions has been significantly reduced compared to Wilkes’ design.
15) The Control field is typically the lone field in a micro-instruction.
16) The Control field specifies the generation of the control signals.
17) There won’t typically be an address field on microinstructions.
18) Instead, after each micro-instruction, pPC will simply be increased.
19) This holds true provided that the p-program is run sequentially.
20) An address will be filed if there is only a branch p-instruction.
21) The branch address will be instantly loaded into CMAR if the branch is unconditional.
22) The Branch condition will verify the proper flag for Conditional branches.
23) A MUX with all flag inputs is used for this.
24) The MUX will instruct CMAR to load the branch address if the condition is true.
25) The CMAR value will merely be increased if the condition is false.
26) Because FLASH ROM is readable but non-volatile, it is typically used to implement the control memory.
ADVANTAGES :
1) Flexibility is the main benefit.
2) The control unit can be changed in any way by just altering the micro-instruction.
3)This makes upgrading and changing the Control Unit relatively simple.
4) In addition, software is much easier to debug than a big Hardwired Control Unit.
5) The majority of micro-instructions don’t require an address field because they are often executed consecutively. 6) As a result, the Control Memory and the size of micro-instructions are both greatly reduced.
DRAWBACKS :
1) The processor must have internal control memory, expanding its size. 2) This pushes up the processor’s price as well.
APPLICATIONS OF MICROPROGRAMMED CONTROL UNIT / MICROPROGRAMMING Microprogramming has a number of benefits, including simplicity, cost effectiveness, and adaptability. It thus has a significant impact on the applications below.
1) CONTROL UNIT DEVELOPMENT:
The instruction sets on modern processors are enormous and intricate. Such processors’ Control Units can be created using microprogramming because they are far less complex and require little maintenance.
2) USER TAILORING OF THE CONTROL UNIT:
The Control Unit can be readily reprogrammed because it was created using software. This can be used to modify the Control Unit specifically. The Control Memory needs to be writable like RAM or Flash ROMs for this to be possible.
3) EMULATION:
Emulation is the technique of instructing one processor (A) to mimic the actions of another processor (B). To accomplish this, “A” must be able to carry out “B’s” commands. The Control Memory of “A” can imitate the actions of “B” for every instruction if the Control Memory of “A” is reprogrammed to be identical to that of “B”. Only microprogrammed control units are capable of doing this. Usually used when a main processor needs to mimic a math co-actions processor.
4) OPERATING SYSTEM IMPROVEMENT:
It is possible to construct intricate and secure OS functionalities via microprogramming. The OS gains more strength and efficiency as a result, but more crucially, it becomes more safe because it is better protected against dangerous virus attacks.
5)HIGH LEVEL OF LANGUAGE SUPPORT:
Many more sophisticated and complex data types are available in modern high level languages. Support for such data types can be provided through microprogramming at the CPU level. The effect is that the language is quicker to execute and easier to compile.
6) MICRO-DIAGNOSTICS:
Debugging a problem with a software-based microprogrammed control unit is much simpler than with a hardwired, complicated control unit. This enables the Control Unit to monitor, identify, isolate, and fix any form of system faults. It may occasionally serve as a runtime replacement if the matching hardwired component malfunctions.
7) IMPLEMENTATION OF SPECIAL PROCESSORS:
Not every processor is a general-purpose device. Specialized processors are needed for many applications, such as DSP (Digital Signal Processors) for communication and GPU (Graphical Processing Unit) for image processing. Microprogrammed Control Units are the best option because they feature intricate instruction sets and require frequent updating.
FORMAT FOR MICRO-INSTRUCTIONS
The control field of the micro-instruction is its primary component. It chooses the control signals that will be generated. It can be in either a horizontal or vertical format.
1)HORIZONTAL MICRO-INSTRUCTIONS:
In this case, each bit of the microinstruction matches a control signal. The micro-instruction will generate that specific control signal if any bit is set to “1”.
2)VERTICAL MICRO—INSTRUCTIONS:
In this case, portions of the micro-instruction must be decoded. The control signal to be produced is determined by the decoded output.
Example of a Micro-instruction 3 There are 3 bits to be deciphered. 3 To Generate A Control Signal 3
| Horizontal Micro-Instructions | Vertical Micro-Instructions |
| Every bit of the micro-instruction corresponds to a control signal. | The production of control signals requires the decoding of bits from the micro-instruction. |
| Does not require a decoder. | requires a decoder. |
| N bits in the micro-instruction will totally produce N control signals. | The microinstruction’s N bits will all be used to create 2″ control signals. |
| Multiple control signals can be produced by one micro-instruction. | One micro-instruction can only produce one control signal. |
| As the control signals increase, the micro- instruction grows wider. Hence the Control Memory grows Horizontally. | More micro-instructions are required to generate additional control signals. The Control Memory increases vertically as a result. |
| Executes faster as no decoding needed. | slower because decoding is required. |
| Micro-instruction are very wide. Hence Control memory is large. | Micro-instructions are a lot more focused. Hence Little control memory is used. |
| Circuit is simpler as a decoder is not needed. | A decoder is required, so the circuit is more complicated. |
MEMORY
MEMORY HIERARCHY
1) Programs and data are stored on memory devices to serve this purpose.
2) The computer uses a variety of memory devices to create a memory hierarchy.
3) Each participant contributes in a certain way to the speed, economy, mobility, etc.
REGISTERS
1) There are registers in the processor.
2) They resemble a pair of flip-flops in most ways.
3) They can directly participate in mathematical and logical activities and store data and addresses.
4) They are often only a few bytes in size, which is incredibly little.
PRIMARY MEMORY
1) It is the primary memory, often known as the initial type of memory.
2) It consists of semiconductor memories made up of RAM and ROM. (Chips as memory)
ROM is not volatile. It is employed for storing permanent data, such as the BIOS software. It normally ranges in size from 2 to 4 MB.
4) RAM is writable and so employed for routine tasks. We load every file into RAM before accessing it from secondary memory. RAM is normally 4 GB to 8 GB in size to offer a lot of working space.
SECONDARY MEMORY
1) Increasing storage capacity affordably is the primary goal of secondary memory.
2) The Hard Disk is the main component. Here is where all of a computer’s files are kept.
3) It is non-volatile and writable.
4) A HD typically has a 1 TB size.
5) Disk memories are far less expensive yet substantially slower than chip memory.
PORTABLE SECONDARY MEMORY
1) These are necessary for the actual file transfers between machines.
2) Floppy Disk: This type of storage uses magnetism. Size on average is 1.44 MB.
3) CD: This optical storage medium is used. The average size is 700 MB.
4) DVD: This optical storage format is used. Size on Average is 4.7 MB.
5) Pen Drives & Memory Cards: This type of storage uses semi-conductors. It has FLASH ROM in it.
It is a unique kind of ROM that is both writable and non-volatile. Usual Size varies based on price from 1 GB to 64 GB.
CACHE MEMORIES
1) Because it employs SRAM, it is the fastest type of memory (Static RAM).
2) DRAM is used in the Primary Memory (Dynamic RAM).
In comparison to DRAM, which utilises capacitors, SRAM is much faster since it employs flip-flops.
4) Yet, SRAM is also significantly more expensive than DRAM.
In order to allow the processor to access the file directly, only the current section of the file that we need to access is copied from Main Memory (DRAM) to Cache memory (SRAM).
6) This provides the highest performance while maintaining a cheap cost.
7) The average cache size ranges from 2 to 8 megabytes. 8) A cache is unified if it contains both code and data; otherwise, it is separated. 9) There are three different sorts of caches, according to their location: LI, L2, and L3. 10) The CPU has a split cache, typically 4–8 KB in size, called 1.1 cache.
MEMORY CHARACTERISTICS
1) Location:
Three different types of memory can be distinguished based on their physical locations.
• On-Chip: The CPU contains this memory. Internal Registers and 1.1 Cache, for instance.
• Internal: The motherboard contains this memory. E.g.:: RAM.
• External: The motherboard is linked to this memory. Like a hard drive.
2) Storage Capacity :
This shows how much info the memory can hold. Of course, it ought to be as big as feasible.
N x M is used to symbolise it.
N here denotes the quantity of memory places (no of words)
M = Number of holes per memory location (word size)
For example:: (4K x 8) denotes 4K locations with 8 bits each.
3) Transfer Modes :
Two methods can be used to access data from memory.
• Word Transfer: In this case, the Processor will only send the data it actually requires.
E.g.:: Data accessed via L1 Cache.
• Block Transfer: In this case, the CPU will transfer a whole block containing the data it needs. This speeds up future access to the block’s remaining data. On the principle of spatial locality, this is based. Near the present place being accessed, a processor is most likely to access data.
4) Access Modes
Data can be obtained from memories in two different ways.
• Serial Access: In this case, places are accessed sequentially, one after the other. The access time varies with the distance between the present location and the target location. The longer the travel time, the farther away the site.
Magnetic tapes, for instance.
• Random Access: Here, you can directly access any location in any order at random. This indicates that regardless of their address, all locations have the same access time.
For instance: RAM, the majority of modern memories.
5) Physical Properties
Memory has a variety of physical characteristics.
• Writeable: The memory’s contents can be changed.
E.g.:: RAM
• Non-Writeable: The memory’s contents cannot be changed.
E.g.:: ROM
• Volatile: When the power is turned off, the memory’s contents are gone.
E.g.:: RAM
• Non-Volatile: When the power is turned off, the memory’s contents remain still there.
E.g.:: ROM The majority of secondary memory, like hard drives, are both writeable and non-volatile.
6) Access Time (tA)
- It is the amount of time between making the request and sending the data.
- It needs to be as little as possible.
- Latency is another name for it.
7) Reliability
- It is the period of time during which the memory is anticipated to keep the data error-free.
- It is measured as MTTF: Mean Time To Failure.
- That ought to be as high as is practical.
8) Cost
- This shows the price associated with storing data in memory.
- Cost/bit is how it is described.
- That has to be as low as it can be.
9) Average Cost
For the overall memory store, it represents the total cost per bit. Think about a device with two memories: M1 (RAM) and M2 (ROM) The average cost of the memory is computed as CAVG = (C1 SI + C2 S2)/(S1 + S2) if C1 is the cost of memory M1 of size Si and C2 is the cost of memory M2 of size S2.
The average cost is decreased by using big sizes of inexpensive memory and small sizes of pricey memory.
10)Hit Ratio (H)
Think about the two memories M1 and M2. M1, such as RAM, is located closer to the processor than M2, such as a hard drive.
A Hit is when the desired data is discovered in M1, and a Miss is when it is not. Let N1 represent the quantity of hits and N2 the quantity of errors. The Hit Ratio H is calculated by dividing the total hits by the total attempts. H = (N1) / (N1 + N2) A percentage is used to express it. H will never be perfect. The majority of computers keep it at around 98%.
As no one memory can serve every requirement, a hierarchy of memories is necessary.
The most expensive and fastest memory are cache memories. Although much slower, hard discs are writeable, non-volatile, and also extremely affordable.
It takes CD/DVD, etc., to be portable. ROM, which is nonvolatile and utilised to store BIOS, is used.
DRAM makes up the majority of main memory since it is readable, faster than a hard drive, and less expensive than SRAM.