Highlights from HPE Discover 2023

At the annual HPE Discover event in Barcelona on Thursday, Hewlett Packard Enterprise (HPE) made several new and enhanced artificial intelligence (AI) and hybrid cloud announcements. These include storage, digital twins solution, and an extended collaboration with Nvidia for AI tuning. The goal is to allow enterprises to optimise generative AI model development and implementation.

“AI is exploding,” said CEO Antonio Neri during the company’s keynote. HPE reported a 37% increase in high-performance computing and AI to $1.2 billion for its fourth fiscal quarter, which ended on October 31.

Neri also mentioned that HPE added $600 million in incremental AI orders in the last few weeks alone. He emphasized that the demand for AI solutions is growing significantly and expects AI orders to accelerate even further in the second half of the new fiscal year, which began on November 1. Neri predicts that 2024 will be the year of AI growth.

File Storage for AI

HPE GreenLake for File Storage, an all-flash data platform, is adapting to provide enhanced support for growing and evolving large-scale AI workloads. This includes those based on generative AI and large language models (LLMs). The platform has been upgraded for AI model training and tuning, with improvements such as increased performance density, integration with Nvidia's new Quantum-2 InfiniBand for better GPU-focused computing, and significantly boosted throughput and capacity.

HPE plans to launch a cloud-based service powered by its supercomputers to support advanced AI applications. Initially focusing on LLMs, HPE intends to extend this service to specialised AI applications in areas such as climate modeling, healthcare, finance, manufacturing, and transportation. These cloud-based services, part of the HPE GreenLake portfolio, will be distinct from other GreenLake offerings as they will be exclusively cloud-based. The solution will be available in the early first half of 2024.

HPE Collaboration with Nvidia for AI Tuning 

HPE and Nvidia are collaborating to develop a computing hardware and software platform specifically for generative AI tuning and inferencing in enterprise data centres. The aim is to make it easier for businesses of all sizes to adapt and use AI. This new offering is designed to be user-friendly, allowing companies to customise AI models with private data and deploy them from local devices to cloud-based systems.

The hardware component of the platform is based on a rack-scale architecture. That said, computing components are managed at the level of an entire rack of servers as a single entity to optimise resources. It includes HPE ProLiant Compute DL380a servers incorporating Nvidia L40S GPUs and Nvidia BlueField-3 data processing units (DPUs), the company said.

GreenLake Momentum Continues

In June 2023, HPE entered the AI cloud market with GreenLake for large language models (LLMs), a public cloud service that enables enterprises to privately train, fine-tune, and deploy large-scale AI using supercomputers capable of running LLMs more efficiently than hyperscale infrastructure.

Regarding GreenLake, HPE reported a fourth-quarter annualised revenue run rate of $1.3 billion, a 37% increase from the previous year. HPE also added 2,000 new GreenLake customers, bringing the total customer base to 29,000 with a total contract value exceeding $13 billion.

Updated Data Lakehouse for Ezmeral

The new AI platform also includes HPE Ezmeral Software, which has been updated to make it easier and quicker for businesses to handle data, analytics, and AI tasks. The software platform allows users to run cloud or on-premise applications in containers and is designed to work smoothly across various cloud environments.

Some of the updates to Ezmeral include a more efficient hybrid data lakehouse, optimised for GPU and CPU usage, facilitating easier data management and analysis. Integration with HPE Machine Learning Development Environment Software improves model training and tuning. The software also features better GPU allocation management, enhancing performance across different workloads.

The company also said that Ezmeral now supports more third-party tools, such as Whylogs for monitoring models and Voltron Data for faster, GPU-accelerated data queries. Customers will be able to order the new AI platform and the various updated components in the first quarter of 2024.

New GreenLake Flex Solution for Digital Twin

HPE also announced a GreenLake Flex Solution for Digital Twin, built on Nvidia OVX-certified HPE ProLiant Gen11 servers with Nvidia L40S GPUs. A digital twin is a virtual model of a real-life object or system. The solution allows users to create, simulate, and optimise virtual assets and processes. It's a scalable, multi-GPU framework that provides the necessary infrastructure, software, and services to leverage the benefits of industrial digitalisation. 

Neri said, the digital twin features AI-accelerated infrastructure and Nvidia Omniverse Enterprise, designed to enable rapid insights from all available data. HPE states that the solution combines the flexibility of a public cloud with the security of a private cloud.

AI and Edge Drive Q4 Performance

These announcements followed HPE's fourth-quarter earnings report, which showed a decline in overall revenue and profit but an increase in high-performance computing, AI, and intelligent edge. While overall revenue declined by 7% year over year to $7.4 billion, and traditional compute sales for servers were down by 31% to $2.6 billion, intelligent edge revenue jumped by 41% to $1.4 billion, and HPC and AI products were up by 37% to $1.2 billion.

[The writer was in Barcelona to attend HPE Discover 2023 hosted by Hewlett-Packard Enterprise (HPE)]