#United States #Santa Clara #AI Inference #d-Matrix #JetStream

d-Matrix Introduces JetStream I/O Accelerators

On September 8, 2025, d-Matrix launched the JetStream, a custom I/O card that promises to revolutionize AI inference capabilities in data centers. This innovation is especially timely, as enterprises are now grappling with a heightened demand for ultra-low latency and effective processing of AI workloads.

As more individuals and organizations integrate AI services into their operations, it becomes increasingly vital to ensure that these services can handle significant user engagement without compromising speed. With the rise of advanced AI functionalities, such as reasoning and multimodal interactivity, there is a decisive shift from merely training AI models to deploying them effectively across diverse applications.

Enhancing AI Performance with JetStream

JetStream is engineered to work seamlessly with d-Matrix's Corsair accelerators and Aviator software, allowing for an accelerated performance boost of up to 10 times faster than conventional GPU solutions. Furthermore, JetStream offers three times better cost-performance and energy efficiency, positioning d-Matrix at the forefront of AI infrastructure solutions. This marks a significant leap in making AI more accessible and efficient for users who demand rapid outputs.

“The introduction of JetStream comes at a critical junction as AI adopts multimodal functionalities, compelling users to seek faster interaction levels,” said Sid Sheth, co-founder and CEO of d-Matrix. He emphasized that the integration of JetStream with the Corsair accelerated platform positions d-Matrix as a standout in the tech landscape, providing scalable solutions tailored for modern challenges.

Technical Specifications and Implementation

JetStream functions as a transparent Network Interface Card (NIC) and streaming solution, optimized explicitly for the Corsair accelerators. It's configured in a standard PCIe format, ensuring compatibility with existing Ethernet switches and allowing for easy deployment in current data infrastructures without necessitating costly upgrades.

The full-height PCIe Gen5 cards are capable of achieving a maximum bandwidth of 400 Gbps, thus supporting extensive data transmission needed for high-scale AI applications. As of now, sample units are available, with production slated to ramp up by the end of the year.

Pioneering the Future of AI Workloads

With its commitment to advancing contemporary AI requirements, d-Matrix is pioneering new approaches to overcome existing barriers in AI chip performance, efficiency, and scalability. The company's focus on Digital In-Memory Computing (DIMC) solutions aims to address speed and interaction limitations in generative AI applications, which have often hindered rapid advancements in this field.

Founded in 2019, d-Matrix has gained the backing of leading investors and strategic partners, including Playground Global and Microsoft's M12 Venture Fund. The diversified investments enable d-Matrix to stay ahead in a rapidly evolving AI landscape by innovating solutions that not only meet current needs but also anticipate future industry trends.

For more insights into the latest developments from d-Matrix and updates on JetStream, enthusiasts and industry professionals can visit d-matrix.ai and follow them on LinkedIn.

Conclusion

The introduction of JetStream represents a pivotal moment in the evolution of AI technology. As dependence on AI continues to grow, having infrastructure capable of supporting vast amounts of data processing while maintaining efficiency is crucial. JetStream does just that, positioning d-Matrix as a leader in the AI landscape, ready to tackle the future with cutting-edge technology and remarkable performance metrics.