Leading cloud services company OVHcloud, one of the world’s largest hosting providers, has announced a full range of cutting-edge serverless solutions, which is a big step forward in its AI strategy. The product launch is a reaction to OVHcloud clients’ growing need for AI solutions that are both economical and easily accessible.
Utilizing its 20 years of hosting infrastructure experience, OVHcloud offers a wide selection of data center GPUs as the foundation of its AI service. The company’s unique vertically integrated approach, which would benefit from a custom water-cooling technology, makes it possible to supply environmentally friendly computational resources. Adding to its portfolio, OVHcloud has launched new Compute GPU instances equipped with 1x to 4x NVIDIA H100 PCIe GPUs, featuring high RAM allocation and local high-performance storage. This is in addition to the existing NVIDIA A100 offerings and the upcoming NVIDIA L40S, L4, and A10 options.
OVHcloud has introduced their new HGR-AI-2 references powered by NVIDIA L40S GPU to meet the needs of customers that want dedicated Bare Metal servers with significant GPU capacity for activities like machine learning, deep learning, or deployment of large language models (LLMs). With an FP8 Transformer Engine and fourth-generation Tensor Cores built into NVIDIA L40S GPUs, these servers - which will be available this spring - would offer strong performance for AI workloads including both training and inferencing.
OVHcloud intends to introduce the Bare Metal SCALE GPU range with three additional references later this year. These are designed to handle applications that employ the NVIDIA L4 Datacenter GPU, such as 3D rendering, video processing, and data inference. With improved AI and video capabilities, the NVIDIA L4, which is built on the NVIDIA Ada Lovelace GPU architecture, would provide a flexible solution for a range of applications.
OVHcloud Data Platform
Alongside these technological resources, OVHcloud has unveiled an end-to-end data solution called the OVHcloud Data Platform, which was unveiled at the OVHcloud Summit last November. With more than 50 connections for data collection, storage, administration, analysis, and visualization, this minimal code platform would make data pipeline management and analytics easier. In addition to supporting use cases like CRM augmentation and faster inventory management, it would guarantee data security and adherence to legal requirements.
AI Notebooks, AI Training, and AI Deploy are a few of the serverless AI products from OVHcloud that are intended to help businesses from concept to implementation. For data scientists and machine learning engineers, these tools would provide an accessible and powerful toolset, backed by high-specification OVHcloud AI infrastructure.
Additionally, the business is working on AI Endpoints, a brand-new serverless solution targeted at integrators and developers. With this tool, users won't need to know anything about API endpoints to access a variety of AI models. With the help of OVHcloud's infrastructure, AI Endpoints will give data protection first priority and provide a testing ground prior to API calls. AI Endpoints offers a wide variety of models at launch, including access to top AI models like BERT, YOLO, or Mistral 7B.
AI App Builder
In addition, OVHcloud is introducing AI App Builder, a fully managed serverless solution for creating AI assistants with generative capabilities. Customers may use this tool to choose LLMs, enter private datasets, and release code that is suitable for use in applications. AI App Builder, which is now accessible in Alpha form, represents a significant advancement in the ease of developing generative AI applications.
Together with integrators and partners, OVHcloud is assisting the ecosystem in acquiring AI capabilities. The business intends to attend the World AI Cannes Festival next month in Cannes, France, while providing webinars.
In conclusion, OVHcloud's entry into the AI space, with an emphasis on serverless solutions and GPU-equipped Bare Metal servers, is a calculated step toward improving the usability and productivity of AI for enterprises.
