NetApp AIPod Mini, with Intel

Keep Reading:  AI without expensive infrastructure is here!   

As organisations look to operationalise AI across business functions, the traditional approach of centralised, GPU-intensive, and often cloud-dependent outcomes can present significant barriers in terms of cost, complexity, and time to value. For many departments exploring practical AI use cases, such as intelligent document processing, fraud detection, or customer sentiment analysis, these challenges hamper adoption and meaningful outcomes.  

CPU-based AI inferencing can offer an immediate, accessible path forward. Modern CPUs are increasingly equipped with advanced instruction sets that enable efficient inferencing for a broad range of AI workloads, without the need for expensive, power-hungry accelerators. This makes them ideal for departmental-level deployment, where AI models can be run locally, embedded into existing infrastructure, and tailored to specific business workflows. By avoiding the bottlenecks of shared centralised platforms, departments gain agility, autonomy, and faster iteration cycles. 

The NetApp AIPod Mini with Intel is a validated design that combines infrastructure and software to deliver a set of use cases, streamlining procurement, deployment, and time to value. It’s exciting to be the only EMEA launch partner for this new solution. Please read on to understand the full value.  

Why did something like this need to be created? 

Both CDW and NetApp understand that getting AI into every organisation, at the departmental level, is a key goal to accelerating enterprise and mid-market adoption. The job of training models is out of scope for most organisations; the use of pre-trained foundation models is the route to cost-effective deployments that deliver a measurable ROI. The perception that any form of AI will require millions and millions needs to be dispelled, and AIPod Mini is a great example of the art of the possible. As with everything in technology, it’s about applying the right solution to the problem. Some challenges will require a team of data scientists and the latest Nvidia Grace Blackwell system. Others demand a more delicate approach, and this is where Intel have excelled at adding AI capabilities into a CPU, allowing the cost and complexity to be collapsed. Stuart from NetApp had the following to say when we talked about why they have brought this solution to market.  

Stuart Oliver – Principal Outbound Solution Product Manager - NetApp 

"We see the AI inferencing market as wide open for disruption and growth, crucially, we believe that Inferencing is where the true AI value is created. More customers are moving directly to the “Value Phase” of AI, meaning they will be buying AI solutions instead of just building models from scratch. We see the AIPod Mini with Intel, as a straight-forward, simple and secure way for our partners to support customers by delivering faster ROI and affordable AI for everyone” 

What is the solution, in simple terms? 

The NetApp AIPod Mini with Intel Solution integrates the Intel Inferencing software stack with tested hardware to streamline the deployment and scaling of AI inferencing workloads. The AIPod Mini is a validated, integrated solution composed of X86-based servers, NetApp A20 high-performance storage appliances, the latest Intel XEON 6 CPU, Arista Ethernet networking, and the OPEA AI framework.  

In practical terms, the solution is a reference architecture composed of the following base components: 

• 2 x Application Servers (2 * Intel® Xeon® 6972P Processor 96-Core 2.40GHz)  

• 2 x Arista Switches (DCS-7280CR3A-32S-R) 

• 1 x NetApp AFF A20  

• 1 x Control server 

The application servers and storage can be scaled as required to meet the demands of multiple use cases or varying numbers of concurrent users. The software layer provides a simple and intuitive front end to connect to data sources of choice and brings pre-trained models for the execution of the chosen outcome.  

Who should be looking at this? 

There are many personas within organisations that should consider the AIPod Mini; a few are outlined below. I believe that most of the key use cases that AI can solve are hidden away in the lines of business of most organisations and are not always known to IT leaders. If you are in an IT role, please share the use cases below with your line of business owners and see if they have challenges, we could solve.  

CIO/CTO 

Looking to prove initial use cases that can underpin a wider AI strategy. Or need to bring POC into a production-ready reference architecture for scaling into the organisation.  

Department Head/Business Unit Leader: 

Focused on improving departmental efficiency, reducing risk, and driving business value through technology. When concerned with the strategic impact and ROI of AI initiatives, looking at a pre-built solution that can bring value at pace will be key.  

IT Manager/IT Operations: 

Who is responsible for the IT infrastructure, security, and compliance within the department or organisation? They are concerned with the ease of integration, scalability, security, and manageability of new technologies. 

Data Scientist/Analyst (Departmental Focus): 

Those who are focused on leveraging data to gain insights and build predictive models relevant to the department's specific needs, and who need reliable access to relevant data. 

Key use cases: 

One of the key advantages of the AIPod Mini is the adaptable nature of leveraging the Intel frameworks. This means we can deploy a suite of prebuilt generative AI outcomes that can be applied across multiple sectors. The following six should highlight the solution's flexibility but certainly should not be considered the full scale of the capability.  

Predictive Maintenance 

An iPod Mini can monitor equipment health and predict failures before they occur, minimising downtime and maintenance costs. An example is smart building systems using IoT sensors for HVAC monitoring. 

Customer Service Optimisation 

AIPod Mini chatbots and intelligent call routing to handle queries efficiently while escalating complex issues to human agents. Example: Automated customer service platforms for telecom companies. 

Retail Inventory Management & Demand Forecasting 

AIPod Mini can analyse historical sales data, seasonal trends, promotions, and external factors such as weather or events to predict future product demand more accurately. This helps retailers stock the right amount of inventory and reduce both overstock and stockouts. 

Quality Control and Defect Detection 

AIPod Mini systems with computer vision identify product defects and anomalies with high precision. This ensures consistent quality, reduces waste, and enhances customer satisfaction. 

Efficient Legal Research 

AIPod Mini research tools quickly analyse vast databases of statutes, case law, and regulations, significantly reducing time spent on traditional research and uncovering insights that might be missed through traditional manual processes. 

Demand Forecasting 

An AIPod Mini analyses sales data, market trends, and external factors to accurately predict demand, reducing overstock and stockouts while optimising inventory levels. 

How do we deliver this solution? 

The solution combines NetApp Storage, Arista Networking, your chosen compute OEM and Intel software (including some NetApp secret sauce) with CDW expertise to integrate and deliver. NetApp and Intel have created validated reference architectures to remove the upfront design burden. Then, by leveraging CDW's deep expertise in each vendor and our Logistics and Technology Centre (LTC), we remove the complexity of combining multiple technologies for the fastest time to inference.  

I am excited that CDW has been chosen as the launch partner for this offer. It aligns with our AI strategy of bringing turnkey offerings to the market that can help organisations Accelerate AI Outcomes. Our goal is to help every organisation leverage the transformative power of Generative AI without it becoming a financial burden. Speaking with NetApp EMEA channel VP Kristian, it was great to get a view of why they chose to launch with our team.  

Kristian Kerr – EMEA Channel VP 

"We have chosen to launch our new AIPod Mini with CDW UK&I as they are the perfect partner to reach those customers who are looking for a simple and secure way to deploy their AI platforms with confidence. NetApp and CDW UK&I have a long standing, trusted and collaborative partnership so it makes sense to develop our AI Inferencing strategy with them.” 

Some final thoughts on the outcomes we believe this solution will bring: 

Contextual Accuracy and Precision: 

• Leverages pre-trained Large Language Models (LLMs) to understand key business nuances, ensuring precise results.
• Even subtle differences in wording or interpretation are handled effectively.

Instant Access to Local Knowledge: 

• Integrates with local and proprietary data repositories.
• Tailor AI models to specific departmental needs with flexible deployment options.

Automated Processes and Efficiency: 

• Perform AI tasks closer to data sources for enhanced privacy, security and reliability.

Cost Savings: 

• Optimised processing power due to RAG’s knowledge graph reduces computational load and operating costs.
• Efficient data handling minimises errors and workload on human resources, leading to significant cost savings.

I look forward to exploring the possibilities of this solution as we move from announcement to GA in Q3.