Clock Gating of Streaming Applications for Power Minimization on FPGA’s

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 01 | Jan 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 259
CLOCK GATING OF STREAMING APPLICATIONS FOR POWER
MINIMIZATION ON FPGA’S
M. Venkata Raghudeep1, K. Anji Babu2
1Student, Dept of ECE, A.K.R.G College of Engineering and Technology, JNTU Kakinada, AP, India
2Assistant Professor, Dept of ECE, A.K.R.G College of Engineering and Technology, JNTU Kakinada, AP, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - This project investigates the reduction of dynamic
power for streaming applications yielded by asynchronous
dataflow designs by using clock gating techniques. Streaming
applications constitute a very broad class of computing
algorithms in areas such as signal processing, digital media
coding, cryptography, video analytics, network routing,
packet processing, etc. This project introduces a set of
techniques that, considering the dynamic streaming behavior
of algorithms, can achieve power savings by selectively
switching off parts of the circuits when they are temporarily
inactive. The techniques being independent from the
semantic of the application can be applied to any application
and can be integrated into the synthesis stage of a high-
level dataflow design flow. Experimental results of at-size
applications synthesized on field-programmable gatearrays
platforms demonstrate power reductions achievable with
no loss in data throughput.
Key Words: Clock gating, Streaming applications, Power
saving, Field programmable gate arrays, Data flow.
1. INTRODUCTION
In the past, the major concernsoftheVLSIdesigner were
area, performance,costandreliability;powerconsiderations
were mostly of only secondary importance. In recent years,
however, this has begun to change and, increasingly, power
is being given comparable weight to area and speed. Several
factors have contributed to this trend. Perhaps the primary
driving factor has been the remarkable success and growth
of the class of personal computing devices (portable
desktops, audio- and video-based multimedia products)and
wirelesscommunicationssystemswhichdemandhigh-speed
computation and complex functionality with low power
consumption.
1.1 Streaming Applications
Many applications in embedded systems perform the
same or similar operations on regular sequences of data,
which can be categorized as streaming applications. For
example, on a smart-phone, one can easily find many
streaming applications that are essential to thefunctionality
of the system, like wireless communication, high-definition
video/audio codes and 3D graphics rendering. A streaming
application usually contains a set of independentprocessing
stages. The structure of a H.264 video codec, which is a
typical streaming application. The different stages or the
codec form a pipeline that is used to encode or decode
videos frames. A frame in a H.264 streamcan beanI-frame,a
P-frame or a B-frame.
The streaming processing structure in these applications
creates lots of optimization opportunities, such as pipelined
parallel execution of differentparts,andexploitingdata level
parallelism in certain stages. More importantly, streaming
applications are among the most performance demanding
applications in embedded systems.The rapidintroductionof
these applications becomes an important driving force for
the development of new embedded technologies. So this
thesis focuses on developing energy efficient architectures
and compilation techniques for streaming applications.
1.2 Clock Gating
Clock gating is a popular technique used in many
synchronous circuits for reducing dynamic power
dissipation. Clock gating saves power by adding more logic
to a circuit to prune the clock tree. Pruningtheclock disables
portions of the circuitry so that the flip-flops in them do not
have to switch states. Switching states consumes power.
When not being switched, the switchingpowerconsumption
goes to zero, and only leakage currents are incurred. Clock
gating is a methodology of turning off the clock for a
particular block when it is not needed and is used by most
SoC designs today as an effective technique to save dynamic
power.
The simplest and most common form of clock gating is
when a logical “AND” function is used to selectively disable
the clock to individual blocks by a control signal, as
illustrated in Figure.1 During synthesis, the tools identify
groups of FFs which share a common enable control signal
and use them to selectively switch off the clocks to those
groups of flops. Both of these clock-gating methods will
eventually introduce physical gates in the clock paths which
control their downstream clocks. These gates could
introduce clock skew and lead to setup and hold-time
violations even when mapped into the SoC. However, this is
compensated for by the clock-tree synthesisandlayouttools
at various stages of the SoC back-end flow.

Fig -1: Block Diagram of Clock Gating
2. PROPOSED SYSTEM
Current FPGA families support differentCGstrategiesand
each manufacturer provides its own IP for managing
these different approaches. The methodology described
here is based on using primitives specific to Xilinx FPGA
architectures. However, this methodology can be modified
to support other FPGA vendors primitives.
The execution of a dataflow program consists of a
sequence of action firings. These firings can be correlated
to each other in a graph-based representation using an
approach called executiontracegraphing(ETG).Thegraphis
an acyclic directed graph where each node represents an
action firing, and a directed arc represents either a data or a
control dependency between two different action firings.
3. IMPLEMENTATION
Algorithm:
1) Step1: Initialize en=1.when F=0,AF=0;
2) Step2: Now queue has empty space. Wait with en=1
until AF=1.
If AF=1 go to next state.
3) Step3: Now CG-ckt ready to give en=0. If AF=0again
go to step2.
Else if AF=1, F=0 be in the same state.
Else F=1, AF=1 go to next state.
4) Step4: Now queue is in full state so,disableclk input
to actor by asserting en=0;
5) Step5: Wait until F=0;
If AF=1, F=1 be in the same state.
Else F=0, AF=1 go to next state;
6) Step6: Now queue is in almost full state en=1;
7) If F=0,AF=0 queue has empty spacethengotostep2;
8) Elseif
F=1,
AF=1
again goto step5.
The controller is implemented as a finite state machine
(FSM) having a clock; a reset; input F, for full; input AF, for
almost full; and output EN, for enable. TheAFinputbecomes
active high when there is only one space left on its FIFO
Queue. Its FSM has five states S ={ INIT , SPACE , AFULL _
DISABLE , FULL , AFULL _ ENABLE } . The controller starts
with the INIT state and maintains the EN output port at
active high until F and AF become active low. The activehigh
EN is maintained during the SPACE state.
As a queue becomes full, the state changes to AFULL _
DISABLE .In this state, the EN output passes to an activelow.
A conservative approach is taken in this state astheBUFGCE
disables the output clock on the high-to-low edge. The clock
enables entering the BUFGCE should be synchronized to the
input clock. Once the queue becomes full, the controller
maintains the EN at active low.
Fig -2: State machine of the clock enabling controller.
The power reduction gain of the aforse
mentioned methodology is evaluated by applying it to a
video decoder design. In the reader can find a variety of
RVC-CAL applications for dataflow programs. One of these
applications is the intra MPEG-4 simpleprofiledecoder. Due
to restrictions on the number of clock buffers in Xilinx
FPGAs, the design selected was re factored to result in 32
actors.
The intra MPEG-4 SP description contains 32
actors and it is 4:2:0 decoder which is separated into eight
processing blocks.
Fig -3: streaming application element (De-blocking filter)
with clock gating.

4. RESULTS:
Table -1: comparison table of proposed and extension
method
PROJECT SLICE
LUTS
SLICE
REGISTERS
LUT-
FFS
DELAY
PROPOSED 368 171 97 8.775ns
EXTENSION 268 155 81 8.775ns
4. CONCLUSION
This project presents a CG methodology applied to
dataflow designs that can be automatically included in the
synthesis stage of an HLS design flow. The application of the
power saving technique is independent from the schematic
of application and does not need any additional step or
effort during the “design” of the application at the
dataflow program level. The CG logic is generated during
the synthesis stage together with the synthesis of the
computational kernels connected via FIFO queues
constituting the dataflow network. Conceivably, these
techniques could be extended to other dataflow methods of
computation. Experimental results are very encouraging:
savings in power dissipation achieved with a slight
increase in control logic without any reduction in
throughput have been achieved. Unsurprisingly, CG is
attractive in situations where the design isnotusedtoitsfull
capacity. In these circumstances CG is a simple, automatic,
and effective way to recover power otherwise lost in “idle”
cycles. As a result, this techniqueisparticularlyinterestingin
applications with dynamically varying performance
requirements, when designing to a particular
performance point is impossible, and when power
consumption is deemed costly. Further investigations into
CG should consider more aggressive control logic,
whereby control is given to each individual actor,
allowing greater flexibility to actor inactivity.
Furthermore, it will be necessary to develop tools
that partition complex applications onto the limited
number of clock domains for more efficient
implementations. Lastly, additional considerations could
be given to controlling clock speed and, possibly, voltage
transitions.
REFERENCES
[1] M. Pedram, “Power minimization in IC design:Principles
and applications,” ACM Trans. Design Autom.
Electron. Syst. , vol. 1, no. 1, pp. 3–56, Jan. 1996. [Online].
Available: http://doi. acm . org/10 . 1145/225871 . 225877
[2] Q. Wu, M. Pedram and X. Wu, “Clock-gating and
its application to low power design of sequential
circuits,” IEEE Trans. Circuits Syst. I, Fundam. Theory
Appl. , vol. 47, no. 3, pp. 415–420, Mar. 2000.
[3] G. E. Tellez, A. Farrahi, and M. Sarrafzadeh, “Activity-
driven clock design for low power circuits,” in IEEE/ACM
Int. Conf. Comput.-Aided Design Dig. Tech. Papers (ICCAD)
San Jose, CA, USA, Nov. 1995, pp. 62–65.
[4] E. A. Lee and A. Sangiovanni-Vincentelli, “Comparing
models of computation,” in Proc. IEEE/ACM Int. Conf.
Comput.-Aided Design , Austin, TX, USA, 1997, pp. 234–241.
[5] G. Kahn, “The semantics of simple language for parallel
programming,” in Proc. IFIP Congr. , Stockholm, Sweden,
1974, pp. 471–475.
[6] E. A. Lee and D. G. Messerschmitt, “Static scheduling of
synchronous data flow programs for digital signal
processing,” IEEE Trans. Comput. , vol. 36, no. 1, pp. 24–35,
Jan. 1987.
[7] E. A. Lee and T. M. Parks, “Dataflow process networks,”
Proc. IEEE, vol. 83, no. 5, pp. 773–801, May 1995.
[8] S. Suhaib, D. Mathaikutty, and S. Shukla, “Dataflow
architectures for GALS,” Electron. Notes Theory. Comput.
Sci. , vol. 200, no. 1, pp. 33–50, 2008.
[9] T.-Y. Wuu and S. B. K. Vrudhula, “Synthesis of
asynchronous systems from data flow specification,”Inf.Sci.
Inst., Univ. Southern California, Los Angeles, CA, USA, Tech.
Rep. ISI/RR-93-366, Dec. 1993.

Clock Gating of Streaming Applications for Power Minimization on FPGA’s

More Related Content

What's hot (18)

Similar to Clock Gating of Streaming Applications for Power Minimization on FPGA’s (20)

More from IRJET Journal (20)

Recently uploaded (20)

Clock Gating of Streaming Applications for Power Minimization on FPGA’s