Mosaic: Breaking the Optics versus Copper Trade-off with a Wide-and-Slow Architecture and MicroLEDs
Kaoutar Benyahya, Ariel Gomez Diaz, Junyi Liu, Vassily Lyutsarev, Marianna Pantouvaki, Kai Shi, Shawn Yohanes Siew, Hitesh Ballani, Thomas Burridge, Daniel Cletheroe, Thomas Karagiannis, Brian Robertson
Abstract
Link technologies in today's data center networks impose a fundamental trade-off between reach, power, and reliability. Copper links are power-efficient and reliable but have very limited reach (< 2 m). Optical links offer longer reach but at the expense of high power consumption and lower reliability. As network speeds increase, this trade-off becomes more pronounced, constraining future scalability.
We introduce MOSAIC, a novel optical link technology that breaks this trade-off. Unlike existing copper and optical links, which rely on a narrow-and-fast architecture with a few high-speed channels, MOSAIC adopts a wide-and-slow design, employing hundreds of parallel low-speed channels. To make this approach practical, MOSAIC uses directly modulated microLEDs instead of lasers, combined with multicore imaging fibers, and replaces complex, power-hungry electronics with a low-power analog backend. MOSAIC achieves 10× the reach of copper, reduces power consumption by up to 68%, and offers 100× higher reliability than today's optical links. We demonstrate an end-to-end MOSAIC prototype with 100 optical channels, each transmitting at 2 Gbps, and show how it scales to 800 Gbps and beyond with a reach of up to 50 m. MOSAIC is protocol-agnostic and seamlessly integrates with existing network infrastructure, providing a practical and scalable solution for future networks.
Ask about this paper
Your agent reads all of it.
Lune indexed this paper to the last equation, along with the top-tier papers that cite it. Ask a question and the answer quotes them.
Your agent calls
Luneget_paper_fulltext
Free to start. No credit card required.
Terminal
Install the CLIlune papers fulltext cd033e83-2ce0-4b2c-b3d7-975501dc0bf4Cited by top-tier papers1
Ask how each one uses itBuilds on3
- Splitwise: Efficient Generative LLM Inference Using Phase SplittingPratyush Patel, Esha Choukse, Chaojie Zhang, Aashaka Shah et al.ISCA 2024 · 282 citations
- SiP-ML: high-bandwidth optical network interconnects for machine learning trainingMehrdad Khani Shirkoohi, Manya Ghobadi, Mohammad Alizadeh, Ziyi Zhu et al.SIGCOMM 2021 · 94 citations
- Rethinking Machine Learning Collective Communication as a Multi-Commodity Flow ProblemXuting Liu, Behnaz Arzani, Siva Kesava Reddy Kakarla, Liangyu Zhao et al.SIGCOMM 2024 · 43 citations
Related papers
- Sirius: A Flat Datacenter Network with Nanosecond Optical SwitchingHitesh Ballani, Paolo Costa, Raphael Behrendt, Daniel Cletheroe et al.SIGCOMM 2020 · 204 citations
- Exploiting Digital Micro-Mirror Devices for Ambient Light CommunicationTalia Xu, Miguel A. Chávez Tapia, Marco ZúñigaNSDI 2022 · 8 citations
- Evaluating Ruche Networks: Physically Scalable, Cost-Effective, Bandwidth-Flexible NoCsDai Cheol Jung, Michael B. TaylorISCA 2025 · 3 citations
- Democratizing Direct-to-Cell Low Earth Orbit Satellite NetworksLixin Liu, Yuanjie Li, Hewu Li, Jiabo Yang et al.NSDI 2024 · 45 citations
- Mosaic Pages: Big TLB Reach with Small PagesKrishnan Gosakan, Jaehyun Han, William Kuszmaul, Ibrahim N. Mubarek et al.ASPLOS 2023 · 21 citations
