A recent report from Nutanix, a leader in hybrid multicloud computing, entitled the ‘Nutanix State of Enterprise AI Report’, highlights a crucial global trend: that the adoption of enterprise AI applications and workloads will spark a new wave of IT infrastructure modernisation initiatives focused on data mobility, security and protection.
And this trend, says Gerhard Breytenbach, head of pre-sales & architecting at Pinnacle, is where Nutanix’s offerings shine. “Using the public cloud for training large language models (LLMs) with publicly available data is logical. However, when proprietary data and intellectual property are involved, bringing AI models and training on-premises becomes essential to maintain data sovereignty and also curb costs.”
As a vendor with technologies that underpin the recent launch of Pinnacle’s AI-focused services division, Nutanix stands out by providing a consistent cloud operating model across all clouds, whether private, public or hybrid environments. This makes it an ideal solution stack for running AI workloads, clarifies Breytenbach.
“Once AI testing and development are completed in the public cloud, businesses can seamlessly transition their AI models and training on-premises without refactoring, running alongside existing non-AI workloads.”
Security and data protection at the core
“Nutanix’s report noted that data security, data quality, scalability and speed of development were respondents’ top considerations related to running AI workloads. Additionally, over 90 percent said that security and reliability are important considerations in their AI strategy. Nutanix’s security measures are embedded from day one,” he continues.
The Nutanix Security Development Lifecycle (SecDL) ensures that security is integrated into every development step. This approach means that Nutanix platforms are secure out of the box, avoiding the need for end-users to undertake additional hardening processes.
Nutanix’s Distributed Storage System further protects against data loss, and its native software-based encryption provides in-transit and at-rest data protection, ensuring that proprietary data remains secure.
Pinnacle and Nutanix: a strategic partnership
Nutanix is a cornerstone brand for Pinnacle, enabling the delivery of enterprise-grade cloud solutions to local organisations of all sizes through valued partners, Breytenbach notes.
“AI can be both costly and complex, but Nutanix’s focus on return on investment (ROI) and simplicity mirrors Pinnacle’s own mission to make advanced technologies accessible to organisations within the Southern African Development Community (SADC).”
Nutanix continues to innovate with solutions that are easy to deploy and operate. Through robust validation processes, the company simplifies the design and deployment of AI solutions, ensuring a streamlined customer experience. Nutanix Validated Designs (NVDs) offer bundled solutions that accelerate and simplify AI deployment, providing comprehensive hardware, software and service components to streamline the process.
Nutanix’s recent collaborations with NVIDIA and Dell further enhance Pinnacle’s AI offerings. “As a hardware-agnostic software vendor, Nutanix enables Pinnacle to supply its full-stack AI platform across a variety of hardware choices, including Nutanix’s Supermicro-based NX range and Dell’s XC range.
“The Nutanix approach really simplifies AI adoption, maintains control, and optimises costs for enterprises,” says Breytenbach.
For more information on the Nutanix solutions available locally through Pinnacle, please contact Gerhard Breytenbach at gerhardb@pinnacle.co.za or +27 (0) 11 265-3000. Alternatively, please visit https://www.pinnacle.co.za/nutanix.
Spotlight on Nutanix GPT-in-a-Box
For organisations looking to implement GPT capabilities while maintaining control over their data and applications, Nutanix GPT-in-a-Box, a turnkey AI solution, is among the standout products available through Pinnacle. This solution includes:
- Nutanix Cloud Platform infrastructure on GPU-enabled server nodes;
- Nutanix Unified Storage for running and fine-tuning GPT models;
- Open-source software for AI workload deployment, including PyTorch and Kubeflow;
- Support for LLMs such as Llama2, Falcon, MPT, and Hugging Face;
- Unified user interface for model management and API endpoints;
- Support for Nvidia AI Enterprise software; and
- Nvidia Tensor Core GPUs for enhanced performance.