AI Inference Platform’s Developer Portal Expands Accessibility

NeuReality has reached a significant milestone with the successful delivery and activation of its 7nm AI inference server-on-a-chip, the NR1 NAPU™. This achievement, along with the complete integration of the NR1 AI hardware and software system in the first quarter, marks a major advancement for the company.

The NR1™ AI Inference offering provides businesses and governments with the ability to utilize new AI training models and existing AI applications without the need to invest heavily in GPUs. Despite the performance improvements in AI accelerators, CPUs continue to be the primary bottleneck in AI Inference, leading to high power consumption and costs that hinder the adoption of AI innovations by many organizations.

Chief R&D officer at NeuReality, Ilan Avital, expressed his satisfaction with the readiness of the NR1 system for customer deployment in Q1 2024. The successful bring-up and integration of the NAPU from TSMC Taiwan within a short timeframe of 90 days showcased the team's remarkable capabilities in handling complex technologies.

The NR1 system's ability to meet functionality and performance requirements across various aspects, including server-on-chip (SOC), IP, and software, positions it well for early customer trials. Cloud service providers, financial services, and healthcare sectors are among the target industries for applications such as computer vision, automatic speech recognition, and natural language processing.

NeuReality's Software Development Kit (SDK) is tailored for high-volume AI workloads in enterprise data centers, offering tools for different compute engines and XPUs. The SDK's optimized partitioning simplifies installation, management, and scalability, providing developers with a more streamlined approach to deploying AI Inference.

The SDK empowers developers with a comprehensive toolchain for accelerating AI pipelines, orchestrating processes, provisioning resources, and managing inference runtimes. This level of flexibility allows developers to tailor their AI deployments to suit the specific requirements of their projects, enhancing overall efficiency and performance.

Recognizing the relatively low global AI adoption rate of 35% and even lower rates in the U.S., NeuReality is committed to breaking down market barriers for mainstream industries. By focusing on reducing power consumption and promoting the use of efficient server configurations for AI Inference, the company aims to make advanced AI technologies more accessible to businesses.

NeuReality's NR1-S™ AI Inference Appliance stands out for its superior performance compared to the Nvidia DGX H100 System. With significantly higher data processing capabilities, lower energy consumption, and a more compact design, the NR1-S™ offers a cost-effective solution without the need for a host CPU.

Investors like Mingu Lee from Cleveland Avenue recognize the transformative potential of NeuReality's AI technologies in revolutionizing industries such as retail, foodtech, and beverage. By leveraging AI for applications like retail analytics, conversational AI, and generative AI-powered solutions, businesses can enhance customer experiences and operational efficiency.

To access the new Software Developer Portal, visit here.