XConn Technologies, a leader in next-generation interconnect solutions, has introduced Apollo 2, a hybrid switch that integrates both Compute Express Link™ 3.1 and PCIe® 6.2 on a single high-density chip. Supporting 64 to 260 lanes, Apollo 2 delivers unmatched flexibility, bandwidth, and efficiency for AI, machine learning, and high-performance computing applications.
With increasing demands for data-intensive AI and HPC workloads, Apollo 2 addresses critical interconnect challenges by improving memory performance, scalability, and power efficiency. This breakthrough innovation redefines next-generation computing by optimizing latency-sensitive workloads and enabling seamless integration across computing architectures.
Apollo 2: Advancing High-Performance Interconnect Technology
1. Hybrid Integration of CXL 3.1 & PCIe 6.2
- Compute Express Link™ (CXL™) 3.1:
- Revolutionizes memory pooling and sharing for AI and HPC environments.
- Improves data transfer efficiency, reducing total cost of ownership (TCO).
- Peripheral Component Interconnect Express® (PCIe®) 6.2:
- Doubles bandwidth compared to PCIe 5.0 while enhancing power efficiency.
- Enables seamless high-speed data movement for next-gen accelerators.
- Hybrid Interconnect Advantage:
- Delivers unmatched compatibility and performance across modern computing platforms.
- Ensures high-bandwidth, low-latency connectivity for heterogeneous architectures.
2. Scalable Performance for AI & HPC
- Configurable from 64 to 260 lanes to support varying workloads.
- High-density chip architecture optimizes interconnect scaling across data centers, cloud, and enterprise environments.
- Advanced memory interconnects improve AI model training speed and efficiency.
- Supports real-time AI inference and HPC workloads with minimal latency.
3. Industry-Leading Efficiency & Interoperability
- Enhanced power efficiency: Reduces energy consumption while maximizing throughput.
- Minimizes integration risks by leveraging Synopsys’ silicon-proven IP solutions.
- Interoperability across modern data center architectures, ensuring seamless adoption.
Industry Perspectives on Apollo 2’s Impact
1. Enabling the Future of AI and HPC Connectivity
“The rapid evolution of AI and HPC applications demands proven interconnect solutions. XConn’s Apollo 2 switch demonstrates the advantages of integrating Synopsys’ standards-based PCIe and CXL IP solutions, optimizing performance, power efficiency, and interoperability.”
— Neeraj Paliwal, SVP, IP Product Management, Synopsys
2. Reducing TCO for AI and HPC Environments
“CXL 3.1 revolutionizes memory pooling and PCIe 6.2 doubles bandwidth, reducing total cost of ownership in AI and HPC environments. With Apollo 2, XConn is driving next-gen connectivity, empowering enterprises and hyperscalers with cutting-edge interconnect solutions.”
— Brandon Hoff, Research Director, IDC
3. A Major Milestone in XConn’s Innovation Roadmap
“Apollo 2 is a significant advancement in our technology roadmap. It not only meets the latest standards of CXL and PCIe but also provides unprecedented density and scalability for modern data centers and computing architectures.”
— Gerry Fan, CEO, XConn Technologies
The Future of AI and HPC with XConn Apollo 2
With AI and HPC workloads growing exponentially, traditional interconnect architectures face scalability and efficiency challenges. The Apollo 2 switch sets a new benchmark by integrating CXL 3.1 and PCIe 6.2 in a single scalable, high-performance interconnect solution.
As data centers, cloud computing, and AI-driven enterprises demand higher bandwidth and lower latency, Apollo 2 future-proofs computing environments by delivering the next evolution of interconnect technology.
XConn’s Apollo 2 hybrid switch is a game-changer for AI, HPC, and next-generation computing. By integrating CXL 3.1 and PCIe 6.2 into a single, scalable interconnect solution, it enhances data movement efficiency, reduces power consumption, and unlocks new levels of AI and HPC performance.
With unmatched flexibility and scalability, Apollo 2 is paving the way for the future of high-performance computing, enabling AI, ML, and cloud-driven applications to scale efficiently.
