HPE has announced, by Nvidiathe development of several Solutions for business and agthic. These solutions offer facilities to companies that display models of generative, autonomous and physical. Nvidia AI Computing by HPE offers optimized solutions for higher performance, energy efficiency, security and new capacities, with the aim of offering a private full stack cloud of the list to work. In this way, companies of all sizes can use these solutions to train, adjust and implement their AI models.
Therefore, HPE has expanded HPE Private Cloud AI to provide it with compatibility with the NVIDIA AI Data platform. HPE Private Cloud AI offers a self -service cloud experience by HPE Greenlake Cloud. On the basis of NVIDIA AI Data Platform transforms the data into actionable intelligence, for which it uses a continuous process that uses the accelerated computing of NVIDIA, its network technology, AI software and business level storage. All this integrated in HPE Private Cloud AI.
This platform is designed to expedite the implementation of architectures and models that support Nvidia AI Data Platform, such as Nvidia AI-Q projects and V Nim microservices for Nvidia models call Nemotron, with advanced reasoning abilities.
Additional updates for HPE Private Cloud AI include a new instantaneous development system, driven by NVIDIA accelerated computing and including a control node, end -to -end A -Asa software and 32 TB of integrated storage. In addition, it offers unified and fluid access from the EDGE to the cloud.
Data Fabric Software It acts as the base of the Hpe Private Cloud AI data lake, offering a unified data layer. HPE FABRIC also guarantees that the AI models receive data, both structured and unstructured, in quality streaming, and optimized in hybrid cloud environments.
Private cloud ai now admits the rapid deployment of NVIDIA projectswhich allows to have the AI Applications Library of NVIDIA. Thanks to already validated projects, such as Multimodal PDF Data Extraction and Digital Twins, Hpe Private Cloud AI simplifies the workloads of AI.
OPSRAMP AND AU SERVICES
OpsRamp It now offers advanced GPU optimization capabilities. Among them, the visibility of native AI software batteries. In this way there is an integrated supervision of the performance in training and inference workloads executed in large clusters of accelerated computing of NVIDIA. The new GPU optimization capacity is available through Private Cloud AI, as well as independently, to extend it to large clusters.
These capabilities will also be available as an operation service for AI, combining HPE Complete Care Service with NVIDIA GPU optimization capabilities. Thus, IT teams can improve in problem management and hybrid environments.
As for the new cases of use and services of the agriculture, HPE will be the first client to use Zora Ai for Finance of Deloitte, an autonomous the AI platform that converts static financial reports into dynamic and interactive experiences, on demand. It can be used in financial statements, modeling of scenarios and market studies and competition.
Crewai, which allows companies to develop autonomous solutions to improve efficiency, adaptability and decision -making, binds to the Unleash AI program to offer multiagente automation. Together with Private Cloud AI, it allows to display and climb automation driven by safe agents and adapted to specific company needs.
On the other hand, HPE has presented various professional services for autonomous, designed to identify, develop and deploy autonomous models in several business processes, accelerating its application with NVIDIA NIM and NVIDIA NEMO Microservices, both integrated into the Nvidia Ai Enterprise software platform with Private Cloud AI.
HPE AI SERVERS A NVIDIA BLACKWELL ULTE Y BLACKWELL ARCHITECTURE
Nvidia ai computing by hpe supplies AI servers that admit the entire spectrum of training, adjustment and inference of AI models Thanks to the platforms Blackwell Ultra and Balkwell from Nvidia. These AI servers can be implemented with accelerated computing of NVIDIA, Networks and the AI Enterprise of Nvidia.
NVIDIA GB300 NVL72 By HPE will allow service providers and companies deploying clusters capable of training models with billions of parameters, also using the HPE experience in liquid cooling. NVIDIA GB300 NVL72 offers high performance with optimized omputation, greater memory capacity and high performance networks for reasoning capabilities in artificial intelligence. This AI is also based on agents and video inference.
The servers HPE ProLiant Compute XD It will be compatible with the NVIDIA HGX B300 platform, designed for equipment that allows training, adjusting and executing large -scale AI models for complex work loads, including autonomous and reasoning in inference in time of proof.
HPE ProLiant Compute DL384b Gen12 Incorporates the superchip Nvidia GB200 Grace Blackwell NVL4, designed for combined high -performance computing work loads and IA. Meanwhile, the HPE ProLiant Compute DL380a Gen12 Incorporates the NVIDIA RTX Pro 6000 Blackwell Server Edition. It is a PCIE -based solution for data centers prepared for inference tasks of the business and visual computing.
The last generation HPE ProLiant Compute Gen12with protection of all phases of the Life cycle of the HPE ILO service, integrates an advanced security processor. This, known as Secure Enclave, generates an unwavering confidence chain against firmware attacks and guarantees complete traceability from the factory to the entire HPE supply chain. In addition, HPE Ilo 7 offers the first server with post-charting, designed to meet the requirements of the cryptographic safety standard of the FIPS 140-3 level 3 certification.
An intelligent and unified layer for AI in HPE
HPE has also announced a unified dates for. Thus, through Greenlake Cloud, companies will be able to obtain information with more agility already through simplified access, with integrated intelligence and a governed data control in multicloud and multiproveing environments.
This new layer integrates structured and unstructured data, and is optimized for thanks to the combination of HPE high performance infrastructure and its advanced business storage, with data intelligence. By deploying with the NVIDIA AI data platform, it allows the food, models and agents of AI intelligently, with data prepared for AI.
The HPE Data Fabric software, which will now be compatible with Alletra Storage MP X10000 and Apache Iceberg, will act as the backbone of the Data Lake Private Cloud and AI and the unified data layer. HPE Data Fabric deals that the models of AI continuously receive structured, unstructured and streaming, optimized and high quality data.
As we have seen, HPE will support the NVIDIA AI data platform through HPE Private Cloud AI. He will also do it with HPE ALLETRA STORAGE MP B10000, ALLETRA STORAGE MP X10000 AND GREENLAKE FOR FILE STORAGE.
This platform will optimize the IA reasoning workloads with specialized artificial intelligence agents that use NVIDIA AI Enterprise software. HPE ALLETRA STORAGE MP will also expand the support for NVIDIA Networking. On the other hand, HPE Greenlake for File Storage, file storage service for private cloud AI, is validated for business infrastructure.
The improvements in HPE ALLETRA STORAGE MP X10000 will allow the generation of objects of objects ready for AI through a new automatic labeling of online metadata. Thus, the data can have intelligence from the moment of storage, accelerating their use by AI applications.
To simplify data management, Alletra Storage NP B10000 has expanded to allow more diverse workloads to address. For this, it now offers unified access to files, an improvement in ransomware protection and the possibility of moving data more easily between local data centers and public clouds.
HPE Alletra Block Storage for Azure It facilitates data management and the distribution of workloads in the hybrid cloud by storage defined by software in Microsoft Azure. In addition, the combination of native detection and recovery of the B10000 matrix with Zerto improves data safety in all storage environments and applications.
