At Vultr, we’re committed to pushing the boundaries of cloud innovation. Our latest partnership with AMD, Broadcom, and Juniper Networks represents a significant step forward in cloud infrastructure. Together, we’re introducing state-of-the-art data center architecture at our expanded Chicago data center region and our new Centersquare facility.
Building an open ecosystem for AI
This collaboration combines industry-leading technologies to create a robust, scalable, and flexible cloud environment optimized for AI workloads. This infrastructure's core is the AMD Instinct™ MI300X GPU accelerator, designed for high-performance training and inference. Paired with the ROCm™ open software platform, these GPUs provide the computational power to handle today’s most complex AI applications.
Broadcom’s advanced Ethernet network adapters ensure secure, high-speed connectivity to support these capabilities. Their technology is tailored for data center environments, enabling intelligent data flow and seamless operation of AI models. Complementing this, Juniper Networks provides AI-native networking solutions that deliver robust, secure, and sustainable connectivity from the data center to the cloud edge.
Why this partnership matters
The demand for powerful AI infrastructure continues to grow, and enterprises need solutions that can scale with their ambitions. This new supercompute cluster offers the performance, reliability, and flexibility required to effectively train, deploy, and scale AI models. By integrating AMD, Broadcom, and Juniper technologies, we’re creating an open ecosystem that supports innovation and accelerates time-to-insight for businesses worldwide.
This development builds on our ongoing collaboration with AMD to bring cutting-edge GPU solutions to Vultr’s cloud platform. Partnering with Broadcom and Juniper Networks extends this vision, ensuring that our infrastructure can meet the evolving demands of AI development – from the data center to the edge.
Unlocking new possibilities with AMD Instinct™ MI300X GPUs
Vultr is also excited to announce the general availability of AMD Instinct™ MI300X accelerators on the Vultr Cloud Platform. With thousands of MI300X GPUs available, clusters of any size can be deployed for reliable, high-performance computing. These GPUs offer best-in-class memory capacity and bandwidth, enabling larger models and datasets to be processed efficiently in-memory.
Optimized for AI training, inference, and HPC workloads, the MI300X GPUs are powered by the open AMD ROCm™ software ecosystem, eliminating concerns about proprietary software lock-in. Transparent, predictable pricing starts at $1.841/hour.
These GPUs address diverse industry use cases:
- Accelerating fraud detection in financial services
- Enabling genomic research in healthcare
- Optimizing manufacturing processes
- Enhancing telecommunications networks
- Supporting high-end CGI rendering
- Powering predictive analytics in retail
Setting a new standard for AI infrastructure
Our expanded Chicago data center region is more than just the latest data center from Vultr featuring large-scale GPU clusters – it’s a launchpad for the future of AI. By combining powerful hardware, advanced networking, and open software, we provide the tools enterprises need to drive breakthroughs in AI. This architecture sets a new standard for cloud infrastructure, offering the performance and scalability necessary to unlock new possibilities in artificial intelligence.