Sign up to save your podcasts
Or
Share Dragonfly Topology for Scalable AI Networks
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
Dragonfly Topology for Scalable AI Networks
This episode explores the 2008 Dragonfly network topology paper and why its ideas suddenly matter again for large-scale AI systems in 2026. It explains how Dragonfly uses high-radix routers and router groups to keep most traffic to a local hop, a single global hop, and another local hop, reducing the number of expensive long-distance optical links compared with flattened butterfly and folded Clos designs. The discussion highlights the paper’s core argument that topology and routing must be co-designed around pin bandwidth, cable cost, power, and congestion, with the authors claiming roughly 20 percent lower cost than flattened butterfly and 52 percent lower cost than folded Clos beyond 16K nodes under their assumptions. Listeners would find it interesting because it connects an old supercomputing interconnect idea to modern TPU fabrics, mixture-of-experts traffic, all-to-all communication, and the growing reality that network design now directly shapes AI system performance.
Sources:
1. Dragonfly Topology for Scalable AI Networks
https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/34926.pdf
2. Technology-Driven, Highly-Scalable Dragonfly Topology — John Kim, William J. Dally, Steve Scott, Dennis Abts, 2008
https://scholar.google.com/scholar?q=Technology-Driven%2C+Highly-Scalable+Dragonfly+Topology
3. Flattened Butterfly: A Cost-Efficient Topology for High-Radix Networks — John Kim, William J. Dally, Dennis Abts, 2007
https://scholar.google.com/scholar?q=Flattened+Butterfly%3A+A+Cost-Efficient+Topology+for+High-Radix+Networks
4. Topological Characterization of Hamming and Dragonfly Networks and Its Implications on Routing — Cristobal Camarero, Enrique Vallejo, Ramon Beivide, 2014
https://scholar.google.com/scholar?q=Topological+Characterization+of+Hamming+and+Dragonfly+Networks+and+Its+Implications+on+Routing
5. Slim Fly: A Cost Effective Low-Diameter Network Topology — Maciej Besta, Torsten Hoefler, 2014
https://scholar.google.com/scholar?q=Slim+Fly%3A+A+Cost+Effective+Low-Diameter+Network+Topology
6. Microarchitecture of a High-Radix Router — John Kim, William J. Dally, Brian Towles, Amit K. Gupta, 2005
https://scholar.google.com/scholar?q=Microarchitecture+of+a+High-Radix+Router
7. The BlackWidow High-Radix Clos Network — Steve Scott, Dennis Abts, John Kim, William J. Dally, 2006
https://scholar.google.com/scholar?q=The+BlackWidow+High-Radix+Clos+Network
8. Scalable High-Radix Router Microarchitecture Using a Network Switch Organization — Jung Ho Ahn, Young Hoon Son, John Kim, 2013
https://scholar.google.com/scholar?q=Scalable+High-Radix+Router+Microarchitecture+Using+a+Network+Switch+Organization
9. A Scheme for Fast Parallel Communication — L. G. Valiant, 1982
https://scholar.google.com/scholar?q=A+Scheme+for+Fast+Parallel+Communication
10. Indirect Adaptive Routing on Large Scale Interconnection Networks — Nan Jiang, John Kim, William J. Dally, 2009
https://scholar.google.com/scholar?q=Indirect+Adaptive+Routing+on+Large+Scale+Interconnection+Networks
11. Rationale and Challenges for Optical Interconnects to Electronic Chips — David A. B. Miller, 2000
https://scholar.google.com/scholar?q=Rationale+and+Challenges+for+Optical+Interconnects+to+Electronic+Chips
12. Optical Interconnects for High-Performance Computing — Marc A. Taubenblatt, 2012
https://scholar.google.com/scholar?q=Optical+Interconnects+for+High-Performance+Computing
13. Optical Interconnects for Extreme Scale Computing Systems — Sebastien Rumley, Meisam Bahadori, Robert Polster, Simon D. Hammond, David M. Calhoun, Ke Wen, Arun Rodrigues, Keren Bergman, 2017
https://scholar.google.com/scholar?q=Optical+Interconnects+for+Extreme+Scale+Computing+Systems
14. Mission Apollo: Landing Optical Circuit Switching at Datacenter Scale — Ryohei Urata, Hong Liu, Kevin Yasumura, Erji Mao, Jill Berger, Xiang Zhou, Cedric Lam, Roy Bannon, Darren Hutchinson, Daniel Nelson, Leon Poutievski, Arjun Singh, Joon Ong, Amin Vahdat, 2022
https://scholar.google.com/scholar?q=Mission+Apollo%3A+Landing+Optical+Circuit+Switching+at+Datacenter+Scale
15. Adaptive Routing in High-Radix Clos Network — John Kim, William J. Dally, Dennis Abts, 2006
https://doi.org/10.1145/1188455.1188552
16. Efficient All-to-All Collective Communication Schedules for Direct-Connect Topologies — Prithwish Basu, Liangyu Zhao, Jason Fantl, Siddharth Pal, Arvind Krishnamurthy, Joud Khoury, 2024
https://doi.org/10.1145/3625549.3658656
17. Toward lower-diameter large-scale HPC and data center networks with co-packaged optics — Pavlos Maniotis, Laurent Schares, Benjamin G. Lee, Marc A. Taubenblatt, Daniel M. Kuchta, 2021
https://scholar.google.com/scholar?q=Toward+lower-diameter+large-scale+HPC+and+data+center+networks+with+co-packaged+optics
18. Toward higher-radix switches with co-packaged optics for improved network locality in data center and HPC networks [Invited] — Pavlos Maniotis, Laurent Schares, Daniel M. Kuchta, Bengi Karacali, 2022
https://scholar.google.com/scholar?q=Toward+higher-radix+switches+with+co-packaged+optics+for+improved+network+locality+in+data+center+and+HPC+networks+%5BInvited%5D
19. Exploring the benefits of using co-packaged optics in data center and AI supercomputer networks: a simulation-based analysis [Invited] — Pavlos Maniotis, Daniel M. Kuchta, 2024
https://scholar.google.com/scholar?q=Exploring+the+benefits+of+using+co-packaged+optics+in+data+center+and+AI+supercomputer+networks%3A+a+simulation-based+analysis+%5BInvited%5D
20. Enhanced UGAL Routing Schemes for Dragonfly Networks — Ram Sharan Chaulagain, Xin Yuan, 2024
https://scholar.google.com/scholar?q=Enhanced+UGAL+Routing+Schemes+for+Dragonfly+Networks
21. On Selection Functions in Adaptive Routing — Alejandro Cano, Cristobal Camarero, Carmen Martinez, 2025
https://scholar.google.com/scholar?q=On+Selection+Functions+in+Adaptive+Routing
22. Co-packaged optics (CPO): status, challenges, and solutions — Min Tan and coauthors, 2023
https://scholar.google.com/scholar?q=Co-packaged+optics+%28CPO%29%3A+status%2C+challenges%2C+and+solutions
23. AI Post Transformers: Computation-Bandwidth-Memory Trade-offs for AI Infrastructure — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-04-09-computation-bandwidth-memory-trade-offs-a83f2b.mp3
24. AI Post Transformers: FengHuang for Rack-Scale LLM Inference Memory — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-04-12-fenghuang-for-rack-scale-llm-inference-m-62708e.mp3
25. AI Post Transformers: Serving MoE Models with Disaggregated Expert Parallelism — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-05-19-serving-moe-models-with-disaggregated-ex-6979d2.mp3
26. AI Post Transformers: Lossless Sparse Deltas for RL Networks — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-05-15-lossless-sparse-deltas-for-rl-networks-84d676.mp3
View all episodes
By mcgrofThis episode explores the 2008 Dragonfly network topology paper and why its ideas suddenly matter again for large-scale AI systems in 2026. It explains how Dragonfly uses high-radix routers and router groups to keep most traffic to a local hop, a single global hop, and another local hop, reducing the number of expensive long-distance optical links compared with flattened butterfly and folded Clos designs. The discussion highlights the paper’s core argument that topology and routing must be co-designed around pin bandwidth, cable cost, power, and congestion, with the authors claiming roughly 20 percent lower cost than flattened butterfly and 52 percent lower cost than folded Clos beyond 16K nodes under their assumptions. Listeners would find it interesting because it connects an old supercomputing interconnect idea to modern TPU fabrics, mixture-of-experts traffic, all-to-all communication, and the growing reality that network design now directly shapes AI system performance.
Sources:
1. Dragonfly Topology for Scalable AI Networks
https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/34926.pdf
2. Technology-Driven, Highly-Scalable Dragonfly Topology — John Kim, William J. Dally, Steve Scott, Dennis Abts, 2008
https://scholar.google.com/scholar?q=Technology-Driven%2C+Highly-Scalable+Dragonfly+Topology
3. Flattened Butterfly: A Cost-Efficient Topology for High-Radix Networks — John Kim, William J. Dally, Dennis Abts, 2007
https://scholar.google.com/scholar?q=Flattened+Butterfly%3A+A+Cost-Efficient+Topology+for+High-Radix+Networks
4. Topological Characterization of Hamming and Dragonfly Networks and Its Implications on Routing — Cristobal Camarero, Enrique Vallejo, Ramon Beivide, 2014
https://scholar.google.com/scholar?q=Topological+Characterization+of+Hamming+and+Dragonfly+Networks+and+Its+Implications+on+Routing
5. Slim Fly: A Cost Effective Low-Diameter Network Topology — Maciej Besta, Torsten Hoefler, 2014
https://scholar.google.com/scholar?q=Slim+Fly%3A+A+Cost+Effective+Low-Diameter+Network+Topology
6. Microarchitecture of a High-Radix Router — John Kim, William J. Dally, Brian Towles, Amit K. Gupta, 2005
https://scholar.google.com/scholar?q=Microarchitecture+of+a+High-Radix+Router
7. The BlackWidow High-Radix Clos Network — Steve Scott, Dennis Abts, John Kim, William J. Dally, 2006
https://scholar.google.com/scholar?q=The+BlackWidow+High-Radix+Clos+Network
8. Scalable High-Radix Router Microarchitecture Using a Network Switch Organization — Jung Ho Ahn, Young Hoon Son, John Kim, 2013
https://scholar.google.com/scholar?q=Scalable+High-Radix+Router+Microarchitecture+Using+a+Network+Switch+Organization
9. A Scheme for Fast Parallel Communication — L. G. Valiant, 1982
https://scholar.google.com/scholar?q=A+Scheme+for+Fast+Parallel+Communication
10. Indirect Adaptive Routing on Large Scale Interconnection Networks — Nan Jiang, John Kim, William J. Dally, 2009
https://scholar.google.com/scholar?q=Indirect+Adaptive+Routing+on+Large+Scale+Interconnection+Networks
11. Rationale and Challenges for Optical Interconnects to Electronic Chips — David A. B. Miller, 2000
https://scholar.google.com/scholar?q=Rationale+and+Challenges+for+Optical+Interconnects+to+Electronic+Chips
12. Optical Interconnects for High-Performance Computing — Marc A. Taubenblatt, 2012
https://scholar.google.com/scholar?q=Optical+Interconnects+for+High-Performance+Computing
13. Optical Interconnects for Extreme Scale Computing Systems — Sebastien Rumley, Meisam Bahadori, Robert Polster, Simon D. Hammond, David M. Calhoun, Ke Wen, Arun Rodrigues, Keren Bergman, 2017
https://scholar.google.com/scholar?q=Optical+Interconnects+for+Extreme+Scale+Computing+Systems
14. Mission Apollo: Landing Optical Circuit Switching at Datacenter Scale — Ryohei Urata, Hong Liu, Kevin Yasumura, Erji Mao, Jill Berger, Xiang Zhou, Cedric Lam, Roy Bannon, Darren Hutchinson, Daniel Nelson, Leon Poutievski, Arjun Singh, Joon Ong, Amin Vahdat, 2022
https://scholar.google.com/scholar?q=Mission+Apollo%3A+Landing+Optical+Circuit+Switching+at+Datacenter+Scale
15. Adaptive Routing in High-Radix Clos Network — John Kim, William J. Dally, Dennis Abts, 2006
https://doi.org/10.1145/1188455.1188552
16. Efficient All-to-All Collective Communication Schedules for Direct-Connect Topologies — Prithwish Basu, Liangyu Zhao, Jason Fantl, Siddharth Pal, Arvind Krishnamurthy, Joud Khoury, 2024
https://doi.org/10.1145/3625549.3658656
17. Toward lower-diameter large-scale HPC and data center networks with co-packaged optics — Pavlos Maniotis, Laurent Schares, Benjamin G. Lee, Marc A. Taubenblatt, Daniel M. Kuchta, 2021
https://scholar.google.com/scholar?q=Toward+lower-diameter+large-scale+HPC+and+data+center+networks+with+co-packaged+optics
18. Toward higher-radix switches with co-packaged optics for improved network locality in data center and HPC networks [Invited] — Pavlos Maniotis, Laurent Schares, Daniel M. Kuchta, Bengi Karacali, 2022
https://scholar.google.com/scholar?q=Toward+higher-radix+switches+with+co-packaged+optics+for+improved+network+locality+in+data+center+and+HPC+networks+%5BInvited%5D
19. Exploring the benefits of using co-packaged optics in data center and AI supercomputer networks: a simulation-based analysis [Invited] — Pavlos Maniotis, Daniel M. Kuchta, 2024
https://scholar.google.com/scholar?q=Exploring+the+benefits+of+using+co-packaged+optics+in+data+center+and+AI+supercomputer+networks%3A+a+simulation-based+analysis+%5BInvited%5D
20. Enhanced UGAL Routing Schemes for Dragonfly Networks — Ram Sharan Chaulagain, Xin Yuan, 2024
https://scholar.google.com/scholar?q=Enhanced+UGAL+Routing+Schemes+for+Dragonfly+Networks
21. On Selection Functions in Adaptive Routing — Alejandro Cano, Cristobal Camarero, Carmen Martinez, 2025
https://scholar.google.com/scholar?q=On+Selection+Functions+in+Adaptive+Routing
22. Co-packaged optics (CPO): status, challenges, and solutions — Min Tan and coauthors, 2023
https://scholar.google.com/scholar?q=Co-packaged+optics+%28CPO%29%3A+status%2C+challenges%2C+and+solutions
23. AI Post Transformers: Computation-Bandwidth-Memory Trade-offs for AI Infrastructure — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-04-09-computation-bandwidth-memory-trade-offs-a83f2b.mp3
24. AI Post Transformers: FengHuang for Rack-Scale LLM Inference Memory — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-04-12-fenghuang-for-rack-scale-llm-inference-m-62708e.mp3
25. AI Post Transformers: Serving MoE Models with Disaggregated Expert Parallelism — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-05-19-serving-moe-models-with-disaggregated-ex-6979d2.mp3
26. AI Post Transformers: Lossless Sparse Deltas for RL Networks — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-05-15-lossless-sparse-deltas-for-rl-networks-84d676.mp3