1 Clockless Logic Montek Singh Tue, Mar 23, 2004

2Outline  Classic static logic pipeline: Sutherland  Classic dynamic logic pipeline: Williams/Horowitz

3 A Classic Asynchronous Dynamic Pipeline Williams and Horowitz’s PS0 pipeline:  Structure  Operation  Performance

4 A Classic Approach: PS0 Pipeline Williams/Horowitz (Stanford U.) [ ]: successfully used in fabricated chips [Stanford ’87] [HAL ’90s] successfully used in fabricated chips [Stanford ’87] [HAL ’90s] Implemented using “ dynamic logic” Processing Block Completion Detector Datain Dataout Stage 1 Stage 2 Stage 3 ack data

5 PS0 Pipeline Stage A PS0 stage consists of dynamic gates and a completion detector: Pull-downnetwork “keeper” PC data inputs data outputs Processing Block CompletionDetector ack

6 Dual-Rail Completion Detector  Combines dual-rail signals  Indicates when all bits are valid (or reset) C Done OR bit 0 OR bit 1 OR bit n  OR together 2 rails per bit  Merge results using “C-element” C-element: if all inputs=1, output  1 if all inputs=1, output  1 if all inputs=0, output  0 if all inputs=0, output  0 else, maintain output value else, maintain output valueC-element: if all inputs=1, output  1 if all inputs=1, output  1 if all inputs=0, output  0 if all inputs=0, output  0 else, maintain output value else, maintain output value

7 Precharge  Evaluate: another 3 events Complete cycle: 6 events indicates “done” PRECHARGE N: when N+1 completes evaluation PRECHARGE N: when N+1 completes evaluation  delete data: after next stage has copied it EVALUATE N: when N+1 completes precharging EVALUATE N: when N+1 completes precharging  accept new data: after next stage is emptied PS0 Protocol evaluates evaluates evaluates indicates “done” precharges 3 Evaluate  Precharge: 3 events N N+1 N+2

8 PS0 Performance Cycle Time =

9 Summary: PSO Pipelining Datapaths are latch-free: dynamic gates themselves provide implicit latches dynamic gates themselves provide implicit latches +: chip area savings +: extremely low latency Data items kept separate by control stage deletes data: only after next stage has copied it stage deletes data: only after next stage has copied it stage accepts new data: only if next stage is empty stage accepts new data: only if next stage is empty è distinct data items always separated by “spacers” Control is extremely simple: each controller = single wire completion detector directly controls previous stage completion detector directly controls previous stage +: chip area savings +: low control overhead

10 Comparison to a Clocked Pipeline How would you design the pipeline if you actually had a clock? 1. Replace handshaking with “magic clocking” each stage gets its own clock each stage gets its own clock successive clocks are slightly skewed successive clocks are slightly skewed  essentially, clocked simulation of asynchronous handshaking! – need multiple clock phases! 2. Use a single clock, but insert latches between stages latches are simple, level-sensitive latches are simple, level-sensitive consecutive stages receive complementary clock signals consecutive stages receive complementary clock signals latch Ck Ck’

11 Comparison … (contd.) Cycle Times?

12 Drawbacks of PSO Pipelining 1. Poor throughput: long cycle time: 6 events per cycle long cycle time: 6 events per cycle data “tokens” are forced far apart in time data “tokens” are forced far apart in time 2. Limited storage capacity: max only 50% of stages can hold distinct tokens max only 50% of stages can hold distinct tokens data tokens must be separated by at least one spacer data tokens must be separated by at least one spacer Our Research Goals: address both issues still maintain very low latency still maintain very low latency