Key Highlights
- Netweb Technologies launched Make in India AI supercomputing systems powered by NVIDIA Grace Blackwell platform.
- Tyrone Camarero Spark introduced as one of the world’s smallest AI supercomputers.
- System delivers around 1 petaflop AI performance with 128GB unified memory in desktop form factor.
- Platform supports AI model inference up to 200 billion parameters and fine-tuning up to 70 billion parameters.
- Netweb unveiled GB200-based Tyrone AI systems designed for large-scale AI training and inference workloads.
- Infrastructure capable of supporting AI models scaling up to 10 trillion parameters.
- Systems designed and manufactured in India to strengthen sovereign AI computing ecosystem.
- Shipping of Tyrone Camarero Spark systems has already begun for select AI organizations in India.
Netweb Technologies India Limited has announced the launch of a new generation of ‘Make in India’ artificial intelligence supercomputing systems powered by NVIDIA’s sovereign AI development platforms, marking a major advancement in India’s high-performance computing ecosystem. The company introduced the Tyrone Camarero GB200 AI Supercomputer along with the compact Tyrone Camarero Spark personal AI compute system, aimed at accelerating AI research, development, and enterprise adoption across India.
The newly launched Tyrone Camarero Spark is positioned as one of the world’s smallest AI supercomputers, delivering petascale AI performance in a desktop-sized form factor. Built on the NVIDIA DGX Spark platform and powered by the NVIDIA GB10 Grace Blackwell Superchip, the system integrates Blackwell GPUs, Grace CPUs, NVIDIA networking, CUDA-X libraries, and the NVIDIA AI software stack. The platform enables developers to run inference workloads on AI models with up to 200 billion parameters and locally fine-tune models of up to 70 billion parameters, supporting advanced AI development use cases.
According to the company, the system delivers nearly one petaflop of AI performance with 128GB of unified memory while operating with significantly lower power consumption compared with earlier generations of AI infrastructure. The platform allows developers and enterprises to build and deploy AI agents and advanced software stacks entirely on-premises, supporting the growing demand for sovereign AI computing capabilities in India.
Netweb also introduced its Tyrone AI Supercomputing systems based on NVIDIA Grace Blackwell architecture, designed to support large-scale AI training and inference workloads. Built on the NVIDIA GB200 NVL4 platform, the system combines multiple Blackwell GPUs connected through NVLink technology and Grace CPUs to deliver enhanced performance for scientific computing, generative AI workloads, and large language model processing.
The company stated that the new infrastructure is capable of supporting AI models scaling up to 10 trillion parameters, enabling real-time large language model inference and compute-intensive applications. The architecture incorporates advanced technologies including next-generation Transformer Engine, high-speed NVLink connectivity, confidential computing security features, and liquid-cooled modular server designs for improved efficiency and scalability.
Netweb highlighted that the systems are designed and manufactured in India, aligning with the government’s Make in India initiative and strengthening domestic sovereign computing infrastructure. The rack-scale deployment integrates multiple GB200 systems within a liquid-cooled environment featuring Grace CPUs, Blackwell GPUs, BlueField-3 DPUs, and high-performance storage to meet next-generation AI data centre requirements.
Shipping of the Tyrone Camarero Spark systems has already begun, with initial units supplied to select AI user organizations in India. The company will also showcase its full Tyrone AI product portfolio, ranging from personal AI computing to advanced data centre solutions, at the India AI Impact Summit in New Delhi.