1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal...

18
1 mbining the strengths of UMIST and e Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer Science University of Manchester [email protected]
  • date post

    22-Dec-2015
  • Category

    Documents

  • view

    219
  • download

    2

Transcript of 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal...

Page 1: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

1Combining the strengths of UMIST andThe Victoria University of Manchester

Asynchronous Signal

Processing Systems

Linda Brackenbury

APT GROUP, Computer Science

University of Manchester

[email protected]

Page 2: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

2Combining the strengths of UMIST andThe Victoria University of Manchester

AgendaAgenda• Why asynchronous?

• Applications suited to asynchronous

• Design examples– DSP design– Viterbi decoder

• Future work

• What have we learnt?

Page 3: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

3Combining the strengths of UMIST andThe Victoria University of Manchester

System TimingSystem Timing

• Synchronous– uses global clock – any state changes occur on clock edge– system states predictable so good tools

• Asynchronous– uses events to control timing– timing is more unpredictable– tool support not as good

Page 4: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

4Combining the strengths of UMIST andThe Victoria University of Manchester

Why Asynchronous?Why Asynchronous?

• No clock generation or distribution– timing uses local handshake signals

• Power only consumed when doing

useful work

• No overhead between idle and active

• Low EMI – switching is spread

Page 5: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

5Combining the strengths of UMIST andThe Victoria University of Manchester

ApplicationsApplications• Async - no help to some applications

– power/performance at full activity similar for synchronous and asynchronous!

• Async good for portable systems– battery size and lifetime is important– workload is highly variable– lots of idle time– low EMI requirement

Page 6: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

6Combining the strengths of UMIST andThe Victoria University of Manchester

Low Power DSPLow Power DSP• GSM chipsets are typically based on

microprocessor + DSP

• DSP performs intensive calculations

• Challenge is to meet required

throughput without excessive power– throughput met with parallelism– area traded for increased speed

Page 7: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

7Combining the strengths of UMIST andThe Victoria University of Manchester

Asynchronous ContributionAsynchronous Contribution

• Design Philosophy– optimize design for typical operation– support design for rarer conditions

• usually at expense of increased operation time

• Simpler logic within processing units– energy and area reduction

Page 8: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

8Combining the strengths of UMIST andThe Victoria University of Manchester

DSP Design Examples 1DSP Design Examples 1• Data dependent adder on critical path

– detect completion of carry path– average carry path only half word length

• Address wrap around in circular buffer– synchronous calculates new and possible

corrected value in parallel - two adders– asynchronous new value only – one adder

• if correction required (rare) this done after

Page 9: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

9Combining the strengths of UMIST andThe Victoria University of Manchester

DSP Design Examples 2DSP Design Examples 2• Register File has eight single-read

single-write ported 32-word banks– efficient parallel access to sequential

registers from 4 Functional Units (typical)

• Request conflicts to same bank rare– broadcast mechanism available– genuine conflicts take 1 read cycle per

request rather than 1 clock cycle each

Page 10: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

10Combining the strengths of UMIST andThe Victoria University of Manchester

Viterbi DecoderViterbi Decoder

• Two data streams transmitted depends on current and previous data

• State transitions of encoder with time can be drawn as a trellis

• Decoder reconstructs trellis

Page 11: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

11Combining the strengths of UMIST andThe Victoria University of Manchester

Asynchronous DecoderAsynchronous Decoder

– Clock only used to input and output data•all internal operation is asynchronous•FIFOs buffer data to meet clock demand

Page 12: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

12Combining the strengths of UMIST andThe Victoria University of Manchester

Branch Metric UnitBranch Metric Unit

• Calculates gap between input symbol and four ideal symbols

Page 13: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

13Combining the strengths of UMIST andThe Victoria University of Manchester

Path Metric UnitPath Metric Unit

j+32

j 2j

2j+1BMa

BMa

BMb

BMb

•add-compare-select operation

previous node metric

next node metric

node node

Page 14: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

14Combining the strengths of UMIST andThe Victoria University of Manchester

Node ArithmeticNode Arithmetic

• Serial arithmetic• Counts events• Unary numbers

– Change of state equals count

– one=1111 two=0001 three=1101 etc.

When smaller count empties merge stops

Page 15: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

15Combining the strengths of UMIST andThe Victoria University of Manchester

History UnitHistory Unit

• Records PMU node winners and global winner over many timeslots

• No error -global winner is child of last winner

• Error – need to reconstruct good path– compute parent of global winner and repeat

ONLY until it agrees with good path– can have many backtraces in parallel– backtraces decoupled from placing data into HU

Page 16: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

16Combining the strengths of UMIST andThe Victoria University of Manchester

History UnitHistory Unit

Page 17: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

17Combining the strengths of UMIST andThe Victoria University of Manchester

Low Power ContributionLow Power Contribution• History Unit

– much smaller– highly concurrent independent operation– computation performed minimised

• Path Metric Unit – most of power– smaller, simple, fast +/- units replace

add-compare-select– idea simple but a lot of control complexity

so dissipated a lot of power!

Page 18: 1 Combining the strengths of UMIST and The Victoria University of Manchester Asynchronous Signal Processing Systems Linda Brackenbury APT GROUP, Computer.

18Combining the strengths of UMIST andThe Victoria University of Manchester

What Have We Learnt?What Have We Learnt?• Asynchronous is advantageous to

some applications• Asynchronous design looks very

different from synchronous design• Get very poor results if just translate

from a synchronous design• Design of asynchronous is harder

– timing and control more complex to design