#USA #Cambridge #Myrtle.ai #Napatech #ML Inference

Myrtle.ai, a trailblazer in the machine learning (ML) inference acceleration landscape, has recently announced its support for VOLLO®, a cutting-edge inference accelerator. Launched with Napatech's NT400D1x SmartNIC series, VOLLO achieves extraordinary latency times, often below one microsecond. This breakthrough is set to transform several industries, particularly where rapid inference is paramount, including financial operations, telecommunications, cybersecurity, and network management.

A Game-Changer in Latency Performance

The realm of machine learning is constantly evolving, with businesses increasingly relying on real-time data processing to drive decisions. VOLLO stands out in this context by providing unmatched ML inference latency, which is critical for applications that require instantaneous reactions to incoming data. With its ability to execute various models—ranging from Long Short-Term Memory (LSTM) networks and Convolutional Neural Networks (CNN) to Multi-Layer Perceptrons (MLP) and decision trees such as Random Forests and Gradient Boosting—VLOO takes the lead in flexibility and speed.

Peter Baldwin, CEO of Myrtle.ai, expressed excitement over this collaboration, stating, "We are thrilled to partner with the world's leading SmartNIC provider to deliver unprecedented low latencies for ML inference. Our clients' demands for rapid and efficient inference continue to grow, and this development enables them to fully leverage VOLLO's superior performance."

Unveiling the Advantages of Using VOLLO

The integration of VOLLO within the Napatech SmartNICs enables organizations to run ML inference directly alongside networking processes. This edge-computing approach leads to various advantages, including:

• - Enhanced Security: Reduced data travel time and improved decision-making capabilities in security applications.
• - Cost Efficiency: Allowing for more robust processing without the need for extensive hardware investments.
• - Increased Profitability: Businesses can leverage predictions and insights in real-time, granting them a competitive advantage.
• - Broader Application Range: From automated trading systems to proactive cybersecurity measures, VOLLO supports a wide spectrum of use cases.

Encouraging Adoption of Machine Learning in Finance

Jarrod J.S. Siket, the product and marketing head at Napatech, shared his enthusiasm regarding the partnership, emphasizing the value that VOLLO can add to financial institutions increasingly adopting ML for automated trading. He commented, "Recognizing that low latency is critical in financial benchmarks, we are excited to provide solutions that meet these evolving needs. The VOLLO compiler is designed to enhance how developers integrate our SmartNICs into their ML applications, thus enriching our service offerings."

Accessible Solutions for Developers

Interested developers looking to access VOLLO's capabilities can download the ML-oriented compiler from vollo.myrtle.ai. Here, they can explore the potential latencies achievable for their specific models leveraging the power of the NT400D1x SmartNIC.

Conclusion

As the demand for efficient ML applications escalates across various domains, Myrtle.ai's innovative approach with VOLLO represents a significant leap towards achieving instantaneous inference. This collaboration not only sets a new standard in latency performance but also paves the way for the seamless integration of machine learning into critical operations, enhancing overall operational efficiency and driving growth in competitive markets.

Myrtle.ai is recognized for its expertise in providing world-class inference accelerators on FPGA platforms and is committed to delivering cutting-edge solutions across various sectors, including fintech, speech processing, and recommendations. With VOLLO, the future of machine learning inference looks brighter than ever, promising improved performance and capabilities for businesses looking to harness the full potential of AI-driven technologies.