Plane-Balanced and Deadlock-Free Adaptive Routing for 3D - - PowerPoint PPT Presentation

▶
plane balanced and deadlock free adaptive routing for 3d
SMART_READER_LITE
LIVE PREVIEW

Plane-Balanced and Deadlock-Free Adaptive Routing for 3D - - PowerPoint PPT Presentation

Feb 22, 2024 113 likes •338 views

Plane-Balanced and Deadlock-Free Adaptive Routing for 3D Networks-on-Chip Presented by: Raed Al-Dujaily Authors: Nizar Dahir, Terrence Mak, Alex Yakovlev, Raed Al-Dujaily and Petros Missailidis Outlines Background 3D NoCs

slide-1
SLIDE 1

Plane-Balanced and Deadlock-Free Adaptive Routing for 3D Networks-on-Chip

Authors: Nizar Dahir, Terrence Mak, Alex Yakovlev, Ra’ed Al-Dujaily and Petros Missailidis Presented by: Ra’ed Al-Dujaily

slide-2
SLIDE 2

Outlines

  • Background

– 3D NoCs – Dynamic Programming Routing in 3D NoCs. – Deadlocks – Turn model for adaptive routing

  • Motivations and Contributions
  • Plane-Balanced 3D Routing

– 3D Odd-Even routing – Balanced OE routing – Degree of Adaptiveness

  • Results
  • Conclusion

1/12/2012 2 NoCArc'12 @ Vancouver-Canda

slide-3
SLIDE 3

Background

  • 3D NoCs
  • Die stacking 3D IC technology and

NoC leads to 3D NoC

  • Advantages:

– Smaller form factor – Lower latency – Higher throughput

  • 3D adaptive routing must be:

– Deadlock free – Balanced adaptiveness

1/12/2012 3 NoCArc'12 @ Vancouver-Canda

slide-4
SLIDE 4

Dynamic Programming Network in 3D NoCs

  • For runtime shortest path

computation.

  • A net of dynamic

programming units (DPU’s).

– multi-source single destination – hard coupled with the router – each unit:

  • gets the costs of the

neighbouring units,

  • propagate the minimum cost

after adding its local cost,

  • cost is defined in terms of the

local router congestion (performance counter).

1/12/2012 4 NoCArc'12 @ Vancouver-Canda

slide-5
SLIDE 5

Deadlocks

  • A situation in which two or more

packets are unable to make progress to their destination because they are waiting for each other to release

  • channels. Thus, neither ever does!
  • Can paralyze network

communications.

  • Strategies to deal with deadlocks are;

– Detection and recovery. – Avoidance (the turn model or virtual channels). – Prevention (circuit switching).

1/12/2012 5 NoCArc'12 @ Vancouver-Canda

slide-6
SLIDE 6

Turn Model for Adaptive Routing

  • Deadlock avoidance using the Turn Model:

– West First – North Last – Negative First

  • Odd-Even routing gives higher and more balances

degree of adaptiveness compared to other deadlock free routing algorithms.

– Restricts locations where certain turns can occur. – Offer more balanced degree of adaptiveness.

even column

  • dd column

1/12/2012 6 NoCArc'12 @ Vancouver-Canda

slide-7
SLIDE 7

Outlines

  • Background

– 3D NoCs – Dynamic Programming Routing in 3D NoCs. – Deadlocks – Turn model for adaptive routing

  • Motivations and Contributions
  • Plane-Balanced 3D Routing

– 3D Odd-Even routing – Balanced OE routing – Degree of Adaptiveness

  • Results
  • Conclusion

1/12/2012 7 NoCArc'12 @ Vancouver-Canda

slide-8
SLIDE 8
  • Motivations

– The original turn model for partial adaptive routing initial proposed to 2D and results in uneven degree of adaptiveness. – No turn model is proposed to utilize 3rd dimension for 3D NoCs.

  • Contributions

– Introducing a new approach for extending 2D mesh partially adaptive routing algorithms to 3D. – Plane-balanced degree of adaptiveness is achieved by applying different rules for different layers. – Evaluation of the proposed method under different traffic scenarios

Motivations and Contributions

1/12/2012 8 NoCArc'12 @ Vancouver-Canda

slide-9
SLIDE 9

Outlines

  • Background

– 3D NoCs – Dynamic Programming Routing in 3D NoCs. – Degree of adaptiveness – Deadlocks – Turn model for adaptive routing

  • Motivations and Contributions
  • Plane-Balanced 3D Routing

– 3D Odd-Even routing – Balanced OE routing – Degree of Adaptiveness

  • Results
  • Conclusion

1/12/2012 9 NoCArc'12 @ Vancouver-Canda

slide-10
SLIDE 10

The 3D Odd-Even routing

  • For the 3D Conventional OE, the

following rules are applied :

– Rule 1:odd column: Packets are not

allowed to take North-West turns nor South-West turns.

– Rule 2: even column: Packets are not

allowed to take East-North turns nor East-South turns.

– Rule 3: Up−xy turns are not allowed in

an even xy-plane, and xy-Down turns are not allowed in an odd xy-plane.

  • dd column

NE ES SW WN NW EN SE WS

even column xy-DOWN UP-xy

1/12/2012 10 NoCArc'12 @ Vancouver-Canda

slide-11
SLIDE 11

Balanced OE Routing

  • Let us define the Modified OE

routing which applies the following rules:

– For even xy-plane,

– Rule 4: in odd row: Packets are not allowed to take West-North turns nor East-North turns, – Rule 5: in even row: Packets are not allowed to take South-West turns nor South-East turns.

– Rule 3 is also applied to constrain entering an leaving xy-planes.

  • Balanced OE uses rules 1 and 2 in

even plane and rules 4 and 5 for

  • dd plane.

even row

  • dd row

NE ES SW WN NW EN SE WS xy-DOWN UP-xy

1/12/2012 11 NoCArc'12 @ Vancouver-Canda

slide-12
SLIDE 12

Degree of Adaptiveness

  • For 3D mesh let:

– source node (xs, ys, zs) – destination node (xd, yd, zd) – dx = |xd−xs|, dy = |yd − ys| and dz = |zd − zs|

  • Degree of adaptiveness for:

– Conventional 3D OE

  • Where h is equal to (dx/2) or (dx-1/2) depending on xs and dx

– Modified 3D OE

  • Where q is equal to (dy/2) or (dy-1)/2 depending on ys and dy.
  • In the proposed Balanced odd-even routing, applying

Conventional OE for for odd layers and Modified OE for even layers will result balanced adaptiveness among the planes

y z x

Conventional OE Balanced OE

1/12/2012 12 NoCArc'12 @ Vancouver-Canda

slide-13
SLIDE 13

Outlines

  • Background

– 3D NoCs – Dynamic Programming Routing in 3D NoCs. – Deadlocks – Turn model for adaptive routing

  • Motivations and Contributions
  • Plane-Balanced 3D Routing

– 3D Odd-Even routing – Balanced OE routing – Degree of Adaptiveness

  • Results
  • Conclusion

1/12/2012 13 NoCArc'12 @ Vancouver-Canda

slide-14
SLIDE 14

Experimental Setup

  • 3D mesh NoC with size of 6×6x4
  • Traffic simulation is performed using a modified version of Noxim.

– The router architecture is modified to support 3D NoCs. – The 2D NoC routing algorithms and traffics are modified to support the 3D NoC routings and traffics.

  • The traffics used in our experiments are; Uniform, Transpose, and

Hotspot.

  • The following routing strategies are compared:

– Odd-Even(buffer): Conventional OE rules are ap-plied (Rule 1,2 and Rule 3 are applied for all planes)with buffer level selection strategy. – Odd-Even(DP): Conventional OE with dynamic programming guided selection strategy to guide packets to the least congested path among the available paths between a source and a destination. – Balanced Odd-Even(DP): The proposed Balanced OE routing in which, in addition to rule 3, rules 1 and 2 are applied in an odd xy-plane and rules 4 and 5 are applied in an even xy-plane. Dynamic programming guided selection strategy is also used in this case.

1/12/2012 14 NoCArc'12 @ Vancouver-Canda

slide-15
SLIDE 15

Performance: Random Traffic

1/12/2012 15 NoCArc'12 @ Vancouver-Canda

0.01 0.011 0.012 0.013 0.014 0.015 0.016 0.017 0.08 0.09 0.1 0.11 0.12 0.13 0.14 Packet injection rate (packet/cycle/node) Throughput( flits/cycle/IP)

Balanced Odd_Even(DP) Odd_Even(DP) Odd_Even(Buffer)

0.01 0.011 0.012 0.013 0.014 0.015 0.016 0.017 10 20 30 40 50 60 70 80 90 100 Packet injection rate (packet/cycle/node) Average delay (cycles)

Balanced Odd-Even(DP) Odd-Even(DP) Odd-Even(Buffer)

slide-16
SLIDE 16

1/12/2012 16 NoCArc'12 @ Vancouver-Canda

Performance: Transpose Traffic

0.01 0.011 0.012 0.013 0.014 0.015 0.016 0.017 0.08 0.09 0.1 0.11 0.12 0.13 0.14 Packet injection rate (packet/cycle/node) Throughput( flits/cycle/IP)

Balanced Odd_Even(DP) Odd_Even(DP) Odd_Even(Buffer)

0.01 0.011 0.012 0.013 0.014 0.015 0.016 0.017 10 20 30 40 50 60 70 80 90 100 Packet injection rate (packet/cycle/node) Average delay(cycles)

Balanced Odd_Even(DP) Odd_Even(DP) Odd_Even(Buffer)

slide-17
SLIDE 17

1/12/2012 17 NoCArc'12 @ Vancouver-Canda

Performance: Hotspot Traffic

0.01 0.011 0.012 0.013 0.014 0.015 0.08 0.09 0.1 0.11 0.12 Packet injection rate (packet/cycle/node) Throughput( flits/cycle/IP)

Balanced Odd_Even(DP) Odd_Even(DP) Odd_Even(Buffer)

0.01 0.011 0.012 0.013 0.014 0.015 10 20 30 40 50 60 70 80 90 100 packet injection rate (packet/cycle/node) average delay(cycles)

Balanced Odd_Even(DP) Odd_Even(DP) Odd_Even(Buffer)

slide-18
SLIDE 18

1/12/2012 18 NoCArc'12 @ Vancouver-Canda

Results Summary

slide-19
SLIDE 19

Outlines

  • Background

– 3D NoCs – Dynamic Programming Routing in 3D NoCs. – Deadlocks – Turn model for adaptive routing

  • Motivations and Contributions
  • Plane-Balanced 3D Routing

– 3D Odd-Even routing – Balanced OE routing – Degree of Adaptiveness

  • Results
  • Conclusion

1/12/2012 19 NoCArc'12 @ Vancouver-Canda

slide-20
SLIDE 20
  • A novel method for extending turn model adaptive routing

algorithms from 2D to 3D NoCs is proposed.

  • The method applies different rules for different layers which results

in different restriction on traffic flow for different layers to achieve 3- D plane-balanced approach with higher degree of adaptiveness is achieved.

  • Path diversity analysis and deadlock freeness of the proposed

method are discussed and compared to the conventional 3D odd- even method.

  • Experimental results show that the proposed balanced odd-even

with DPN can achieve improvement of up to 23.8% compared odd- even with buffer level and 8.3% compared to odd-even with DPN and the improvement is consistent for all the considered traffic types.

1/12/2012 20 NoCArc'12 @ Vancouver-Canda

Conclusion

slide-21
SLIDE 21

Thank you for listening …

1/12/2012 21 NoCArc'12 @ Vancouver-Canda