Ultra-low latency FPGA Framework:
Technical specifications for HFT
Unleash technical prowess with our scalable hybrid solution for optimizing high-frequency trading operations
Hybrid ULL FPGA Framework
specifications and capabilities
Dive into the powerful capabilities of Orthogone’s ULL FPGA Framework designed for
High-Frequency Trading (HFT). This hybrid solution enhances speed by integrating
FPGA hardware with adaptable software, providing superior performance and fast
communication between FPGA logic and external devices. It ensures execution times
in the nanosecond range, making it ideal for low-latency trading environments.
The ULL FPGA Framework development kit (FDK) comprises a complete set of IP cores, a comprehensive end-to-end FPGA development environment, and an extensive software framework.
Scalable hybrid solutions for
high-performance data processing
As your trading strategies evolve, our ULL FPGA Framework grows with you. It provides scalable data processing solutions that handle increasing data volumes without sacrificing speed or efficiency, making it a future-proof investment for your trading operations.
Suite of IP cores included
- Full suite of ULL FPGA IP cores (encrypted source code format)
- ULL 10G MAC/PCS: Ultra-low latency Ethernet MAC/PCS
- ULL 10G TCP/IP: Implements a full TCP/IP stack in FPGA with ultra-low latency performances
- ULL 10G UDP/IP: Implements a full UDP/IP stack in FPGA with ultra-low latency performances
- ULL PCIe DMA controller: Ultra-low latency DMA controller to move data between FPGA and host memory rapidly over PCIe interface
Key features
- Full RTL implementation supporting layers 2, 3, 4 (ARP, IPv4, ICMP, TCP, UDP)
- MAC/PCS
- TCP/IP Offload Engine
- UDP/IP Offload Engine
- PCIe DMA Controller
- Supports AMD UltraScale+ FPGA
- Supports AMD Alveo FPGA platforms
- FPGA simulation and verification environment
- Linux CentOS distribution
- Kernel bypass for optimized data path
- Ultra-low latency C/C++ API with zero-copy optimized path
- Advanced network statistics
- Test and debug applications
- A complete suite of Linux drivers, libraries & API
DELIVERABLES
FPGA Development Kit
- Encrypted FPGA IP cores
- ULL 10G MAC/PCS
- ULL 10G TCP/IP
- ULL 10G UDP/IP
- ULL PCIe CBDMA controller
- FPGA Development and simulation environments
- BSP for AMD x3522pv platform
- Reference design examples
- Test and debug applications for latency/throughput measurements
- Product datasheet and user guides
Software Framework
- Linux user-space library and kernel driver module
- Standard RPM installation package
- Source code (C) provided
- Driver module automatically compiled at installation and on kernel update
- Ultra-low-latency network device interface
- Standard network interface using Linux stack for generic protocols
- DHCP, ICMP, IGMP, etc.
- Ultra-low-latency PCIe DMA channels for FPGA data exchange
- Product documentation
- Tools and example applications
- Application to list framework information from host software and FPGA
- Performance application for validation of latency and throughput
- Source code (C/C++) provided and compiled at package installation
Experience it yourself
Discuss with one of our experts to determine if you qualify for a trial.
Latency performances
High speed, flexibility and scalability
Gain exceptional performance for latency-critical Ethernet applications with the industry-leading Ultra-Low Latency 10G Ethernet MAC and PCS. The core is designed using advanced techniques, resulting in outstanding gate count utilization and latency performance.
Hardware acceleration
Unlock the potential of high-performance and seamless network connectivity with our ULL TCP/IP and UDP/IP Offload Engine. Our products address the growing need for throughput and hardware acceleration in the data center industry, providing network protocol offload for applications such as financial data processing, as well as reprogrammable Smart NICs.
Bidirectional data transfer
Harness the powerful ULL DMA capabilities by seamlessly integrating on-chip applications with host-based applications, enabling seamless data delivery for advanced local processing or integration with other sophisticated applications.
High speed, flexibility and scalability
Gain exceptional performance for latency-critical Ethernet applications with the industry-leading Ultra-Low Latency 10G Ethernet MAC and PCS. The core is designed using advanced techniques, resulting in outstanding gate count utilization and latency performance.
Hardware acceleration
Unlock the potential of high-performance and seamless network connectivity with our ULL TCP/IP and UDP/IP Offload Engine. Our products address the growing need for throughput and hardware acceleration in the data center industry, providing network protocol offload for applications such as financial data processing, as well as reprogrammable Smart NICs.
Bidirectional data transfer
Harness the powerful ULL DMA capabilities by seamlessly integrating on-chip applications with host-based applications, enabling seamless data delivery for advanced local processing or integration with other sophisticated applications.
SOLUTION BRIEF – FPGA Solutions for Ulta-low Latency SmartNIC
Quickly develop the ultimate portfolio of flexible FPGA and software applications
Typical use cases
Hardware and software platforms we support
Orthogone’s ULL FPGA Framework is designed to support a wide range of FPGA and software
platforms, for high-performance, low-latency applications.
Customized FPGA and software frameworks for any platform
At Orthogone, we understand that each customer has unique platform requirements. Whether it’s about enhancing existing platforms or developing a new one, we assure you of a seamless and adaptable integration process that is both fast and effective. Our expertise enables us to adapt quickly to your hardware and software environment, guaranteeing you a customized, high-performance solution tailored to your needs.
Supported FPGA Platforms:
- Current: Xilinx/AMD Alveo ULL Accelerator cards (Reference designs: X3522pv, UL3524).
- Roadmap: Xilinx/AMD Versal FPGA and Accelerator Cards (e.g., Alveo V80 Compute Accelerator).
Supported Software Platforms:
- We support multiple CPU architectures, including Intel Xeon and i9. Our solutions have been rigorously tested on Dell PowerEdge servers and high-performance servers like Blackcore Technologies, ensuring their reliability and performance.
- Future: We’re pleased to announce our future support for AMD Ryzen processors.
Legacy Platforms:
- This offering also includes legacy platforms like the Kintex/Virtex UltraScale+ devices and Napatech NT200A02.
Pricing options
At Orthogone Technologies, we offer a pricing model designed to fit the unique needs of our
customers, ensuring both flexibility and cost-effectiveness. Here’s how our pricing works:
licensing option
licensing option
pricing option
Project-based licensing option
- Best for: Customers needing a license for a single, specific project.
- What you get: Use of our ULL FPGA Framework for one named project.
- Ideal for: Teams focused on a particular development effort, allowing them to leverage our technology for targeted outcomes.
initiatives, our pricing model is designed to support your goals efficiently and effectively.
Support and integration
Scalability and support are crucial for sustained success in the high-frequency trading industry. Orthogone’s products are not only built for today’s demands but are designed to scale as your business grows. Our dedicated support team ensures smooth implementation and ongoing optimization, allowing you to stay ahead in a rapidly evolving market.
Our ecosystem of technological partners
Here are the answers to some frequently asked questions:
Our pricing model is designed to offer flexibility and value tailored to the specific needs of our customers. While we remain competitive within the market, our focus is on providing a pricing structure that offers more than just cost savings:
- Customized solutions: Unlike some competitors, who may offer a one-size-fits-all approach, we provide both project-based and site-based licensing options. This allows you to choose a plan that perfectly aligns with your project scope and scale.
- Scalable pricing: Our per-card pricing model ensures you only pay for your needs. This is particularly advantageous for businesses with varying deployment scales, offering a cost-effective solution that grows with your needs.
- Operational flexibility: We offer annualized licensing, turning what could be a significant upfront capital expense into a manageable operational expense. Many of our competitors may not provide this flexibility.
Our pricing competes with but often surpasses our competitors by offering tailored options that maximize value, efficiency, and flexibility for your specific needs.
