Scale Up with 1.6T at OFC
Join us at OFC 2026 to see production-ready 1.6T and 800G transceivers and high-speed copper solutions powering real AI deployments.
Book a meeting with atop @ ofcOFC BOOTH #601 | March 17-19 | Los Angeles Convention Center, Los Angeles, CA
Live and Static Demos
At OFC 2026, we will demonstrate both optical and copper 1.6T technologies designed for next-generation AI fabrics.
A live 1.6T OSFP DR8 functionality demonstration highlighting LRO architecture, ultra-low latency operation, and optical performance optimized for dense GPU and accelerator clusters.
Alongside this, a static demonstration of our 1.6T OSFP AEC 3m solution will highlight high-speed copper interconnects designed for short-reach, power-efficient AI infrastructure. These demonstrations reflect our focus on deployable, production-ready solutions - not concept hardware.
Alongside this, a static demonstration of our 1.6T OSFP AEC 3m solution will highlight high-speed copper interconnects designed for short-reach, power-efficient AI infrastructure. These demonstrations reflect our focus on deployable, production-ready solutions - not concept hardware.
1.6T in Practice: Live at OFC
1.6T OSFP DR8 LRO - Functional Demo
1.6T OSFP AEC 3m - Static Demo
Production-Ready 1.6T for AI Fabrics
Why 1.6T LRO Matters for AI Infrastructure
ATOP’s 1.6T OSFP DR8 SiPh LRO is engineered to meet the demands of modern AI and hyperscale environments where bandwidth, power efficiency, and latency are critical.
Our hot-pluggable OSFP form factor supports an aggregate bandwidth of 1.6Tbps, using eight lanes of 212.5Gb/s PAM4 on both optical and electrical interfaces. The module is based on a next-generation 3nm DSP architecture and leverages silicon photonics transmitters operating at 1310nm with PIN-based receivers.
Designed for Ethernet and InfiniBand environments, the module supports link lengths up to 500 meters over single-mode fiber, while maintaining power consumption below 16W across the full operating range. Commercial temperature operation from 0°C to 70°C enables deployment across a wide range of data center environments.
This architecture enables extremely low latency operation through an LRO design that removes the receive-side DSP, reducing processing overhead and eliminating GPU wait-states in tightly coupled AI fabrics.
Designed for Ethernet and InfiniBand environments, the module supports link lengths up to 500 meters over single-mode fiber, while maintaining power consumption below 16W across the full operating range. Commercial temperature operation from 0°C to 70°C enables deployment across a wide range of data center environments.
This architecture enables extremely low latency operation through an LRO design that removes the receive-side DSP, reducing processing overhead and eliminating GPU wait-states in tightly coupled AI fabrics.
Designed for AI at Scale
Ultra-Low Latency by Design
Lower Power per Terabit
Predictable Performance at Scale
Why 1.6T LRO Matters for AI Infrastructure
As AI clusters scale, latency consistency and power efficiency become just as critical as raw bandwidth.
Our 1.6T LRO architecture delivers ultra-low latency connectivity without receive-side DSP processing, supporting high-density GPU clusters where every microsecond and watt matters. The result is predictable performance, reduced power draw, and improved efficiency across large AI training and inference environments.
This approach enables higher port density, stable operation under sustained workloads, and faster time-to-compute at scale.
This approach enables higher port density, stable operation under sustained workloads, and faster time-to-compute at scale.
Al-Ready Connectivity, From Edge to Cloud
ATOP delivers end-to-end 800G and 1.6T connectivity solutions designed to support AI infrastructure across the full network lifecycle.
From legacy and installed-base platforms through to next-generation AI fabrics, our portfolio enables customers to upgrade when it makes sense - without forced transitions or supply constraints.
We support both optical transceivers and high-speed copper interconnects across AI training, inference, hyperscale, and high-performance computing environments, delivering high bandwidth, high density, and low latency where it matters most.
From legacy and installed-base platforms through to next-generation AI fabrics, our portfolio enables customers to upgrade when it makes sense - without forced transitions or supply constraints.
We support both optical transceivers and high-speed copper interconnects across AI training, inference, hyperscale, and high-performance computing environments, delivering high bandwidth, high density, and low latency where it matters most.
Powering Sustainable AI Infrastructure
Sustainability is becoming a defining requirement for AI infrastructure at scale.
Our 800G and 1.6T copper solutions support greener data center designs by reducing power consumption, improving signal integrity, and simplifying short-reach deployments.
Optimized copper interconnects help lower overall energy usage and contribute to reduced carbon footprints across AI clusters. Our approach supports customers balancing performance demands with long-term sustainability goals.
Optimized copper interconnects help lower overall energy usage and contribute to reduced carbon footprints across AI clusters. Our approach supports customers balancing performance demands with long-term sustainability goals.
Why It Matters at Scale
Lower Power per Link
Optimized for Short-Reach AI Fabrics
Reduced Infrastructure Footprint
Scalable, Traceable Supply, Built for Growth
ATOP designs and manufactures its solutions with scalability, traceability, and compliance at the core.
Our products are certified, tested, and responsibly sourced, supported by automated and validated manufacturing processes that ensure consistent quality across large production runs. We provide actionable sustainability tools and data to support ESG reporting and long-term infrastructure planning.
With intelligent manufacturing and global production capacity, ATOP delivers agile, reliable supply for large-scale 800G and 1.6T deployments worldwide.
With intelligent manufacturing and global production capacity, ATOP delivers agile, reliable supply for large-scale 800G and 1.6T deployments worldwide.
Who Should Meet ATOP at OFC 2026
We’re meeting with hyperscale operators, AI infrastructure teams, data center architects, telecom and transmission providers, wireless network teams, OEMs, and platform partners evaluating or deploying high-speed optical and copper interconnects for AI-driven and high-capacity networks.
If you’re planning AI fabric upgrades, scaling GPU clusters, expanding transport capacity, or evaluating next-generation 1.6T connectivity, we’d welcome the conversation.
If you’re planning AI fabric upgrades, scaling GPU clusters, expanding transport capacity, or evaluating next-generation 1.6T connectivity, we’d welcome the conversation.
Book a Meeting at the ATOP Stand
Use the form below to book a meeting with ATOP during OFC 2026.