Q461. What is the difference between the clock mesh and clock tree-type distribution system?
Clock mesh distribution system:
- Presence of mesh net, which smooths the arrival time difference from multiple mesh drivers.
- Mesh drivers are connected to mesh net as the multi-driven net.
- It produces much lower clock skew and clock insertion delay.
- The design stages are more.
- Power dissipation is high.
- Complexity in implementation
- More routing resources are required to produce clock meshes.
Clock tree distribution system:
- A mesh net is not required.
- number of design stages is optimal.
- Power dissipation is low.
- The clock tree has a clock source, clock tree cells, clock gating cells, buffers, and load.
- It is best suited for slow clock circuit designs.
Q462. Are clock tree synthesis and clock tree distribution, the same thing?
No, clock tree synthesis and clock tree distribution are not the same things.
The clock tree synthesis is used to design the clock tree distribution system.
It is used to minimize the clock insertion delay as well as clock skew. For
clock tree synthesis, ideal clock arrival times are used, whereas the clock tree
distribution system uses real clock arrival time.
Q463. What is the need for clock gating?
The clock gating means controlling the clock toggling activity. As the clock
drives a lot of elements in a circuit, it consumes a lot of power. The ability to
turn off the clock toggling, when not required, is known as clock gating.
In synchronous circuits, clock gating is used to save dynamic power. An
extra logic circuit is used which disables the unused clock states. In RTL
(register transfer level), the clock gating is commonly used to reduce the
size of the die as well as dynamic power consumption. It does not affect the
functionality of the design. The clock gating prunes the clock tree i.e. disable
the switching of flip-flops. The switching of flip-flop consumes power. By
using clock gating, switching power consumption is zero. The clock gating
technology saves the area of the die. The concept of clock gating is shown in
Figure.
Clock gating functionally requires only an AND or OR gate. The other
input of the AND gate is used to turn off the clock for inactive receivers.
Thus, it is an efficient power-saving technique.
Q464. What is the significance of CRPR in static timing analysis?
In static timing analysis, CRPR stands for clock reconvergence pessimism
removal. The static timing analysis is based on a worst-case analysis. In
setup analysis, it uses the slowest possible launch path and fastest capture
path. If launch and capture share a common path, the worst case of STA
becomes pessimistic as in a common path, fast and slow path cannot happen
simultaneously. The CRPR is the accuracy limitation of STA.
Q465. What is DEF and what is its use?
The DEF file is a designed exchange format. It is used to describe:
- Physical aspects of design such as die size, connectivity, macros, etc.
- Floorplanning information such as standard cells, placement, and routing, etc.
- The physical representation of power and signal routings, pins, etc
Q.466. Explain the term metastability?
In a flip-flop, if the setup and hold time violation takes place, it results in
an unpredictable state known as the metastable or quasi-stable state. It can
cause a system failure in digital devices such as FPGA, ASIC, etc. In the
metastable state, the circuit is unable to settle at either logic 0 or logic 1 within
the stipulated time period, this will further fail the system functionality. It
happens due to the toggling of flip-flop during the clock transition.
So, when setup and hold time is violated and the output of flip-flop
inside the FPGA is unknown or indeterministic, this condition is known as
metastability.
Q.467. What are the effects of metastability?
- If fan-out is high, the circuit will go to a metastable state, the flip-flop will toggle unintentionally. There will be unexpected behavior of the system.
- The circuit will draw excessive current.
- The output will have non-deterministic behavior.
- The output of the clocked pass gate does not charge properly.
- The circuit does not meet timing constraints.
Q.468. What are the reasons for metastability?
- Slow transition timing constraints at the input and output level (rise time and fall time).
- Low VDD
- High parasitic capacitances
- Cross talk.
- If the input is an asynchronous signal
- High clock skew
- Excessive combinational delay
Q.469. How metastability can be avoided or tolerated in a circuit?
If input data meet setup and hold time constraints, the problem of metastability can be reduced to an extent. If the signals are generating from different
clock domains, it is difficult to control metastability.
- The clock period should be precise to avoid delay.
- Add one or more successive synchronizing flip-flops to the synchronizer.
- Use metastable hardened flip-flops.
- Provide the needed settling time.
- Receive each asynchronous signal by clocking it into only one flip-flop.
- Use asynchronous reset.
- Use the metastability filter, but it will increase slack.
Q.470. Can you synchronize between two clock domains?
Yes, two clock domains can be synchronized by using either synchronizer
or asynchronous FIFO (if high performance is required). The asynchronous
FIFO has two separate interfaces, one clock for reading and another clock is
for writing or data extraction purposes.