Z by HP Boost slices up GPUs to speed up AI workloads

HP is trying to make AI development, training, and deployment easier with Z by HP Boost. Instead of deploying several expensive workstations, or trying to share one full workstation across a team, Z by HP Boost gives direct, instant access to the GPU resources of a workstation.

We haven’t heard a ton of details about the service yet. HP says it will be available in early 2025 for the U.S. and U.K., and it’ll share pricing closer to when the service becomes available. It sounds like Z by HP Boost is using local GPU resources, however. Earlier this month, Lenovo announced its TruScale GPUaaS (GPUs as a service), which allows enterprise customers to scale up GPU resources on-demand through a cloud-based, pay-as-you-go model. Z by HP Boost sounds like something different.

Instead of cloud-based GPU resources, Z by HP Boost pools together GPU resources across a network and allows you to share them. Your HP OmniBook X will be able to tap into a fleet of workstations, for example, and in a fully remote fashioon. The on-demand nature of the service means there shouldn’t be a lot of configuring, either. If you need more GPU resources, and they’re available through that network, they’ll be available to you.

HP is targeting the feature at AI data scientists who often need access to huge amounts of GPU compute. The feature comes as part of HP’s AI Studio, which was released in April of this year. The goal of the platform is, according to HP, to bring data, people, and compute together in one platform. The new Z by HP Boost aspect solves the compute side of the equation.

Z by HP Boost will work with the other features available in AI Studio. The platform includes tools such as Z by HP Gen AI Lab, which includes tools for training and deploying Large Language Models (LLM). In addition, HP has partnered with Galileo and Nvidia to provide pretrained models and training data inside of AI Studio.

Although AI Studio is technically out, HP hasn’t opened the service widely to customers yet. The company currently has a wait list for new customers to join AI Studio.