• Huawei unveiled its AI computing platform at MWC 2026.
• It says the system shortens build times and supports 150+ AI models.

At this year’s Mobile World Congress (MWC) in Barcelona, Spain, Huawei introduced a new service aimed at helping companies build and operate the computing infrastructure behind AI services. The company calls it the Intelligent Computing Platform Service Solution, and it is now being offered to customers in the global market for the first time.

The solution is designed to give organisations a stronger computing foundation for running AI workloads. Many businesses struggle with planning, building and operating the infrastructure needed to host large models and AI services. Huawei says its new offering is meant to speed up that work and help companies keep their systems running smoothly.

Shorter build times for big data centres

One of the biggest hurdles for companies adopting AI is setting up the hardware and data centre space required to support heavy computing tasks. Many data centre projects take seven to nine months or more to complete from start to finish. Huawei’s platform aims to shorten that timeline by combining design work with on-site construction support. The company says it uses simulation tools for things like power efficiency, liquid cooling and cabling planning to reduce physical work needed on site. This may compress the typical renovation cycle to four to six months.

This kind of reduction in time could be material for firms that need to spin up computing resources quickly to support growth, experiments with LLMs or new AI services. Shorter build times may also reduce labour and logistics costs, especially in crowded markets where skilled infrastructure workers are in demand.

Tools for cluster setup and performance

Beyond planning and construction, Huawei’s solution includes tighter integration of computing clusters. A computing cluster refers to a group of servers and processors working together to handle a big task. Modern AI workloads often depend on clusters with hundreds or thousands of nodes to deliver the necessary performance.

Huawei says its deployment process can get a large cluster running in around 15 days once physical hardware is in place. The company also provides tools for performance tuning and optimisation, which are meant to help AI models run more efficiently across the system.

Efficiency and stability matter because poorly optimised clusters can waste power or fail under heavy loads. In AI applications, erratic performance can slow down businesses and frustrate users who expect consistent response times.

Model adaptations and expert knowledge

Another area the platform touches on is “model adaptation.” This refers to the work of taking a standard AI model and preparing it to run well on specific hardware or within a particular enterprise environment.

Huawei says it has adapted more than 150 mainstream AI models, covering around 90% of common enterprise use cases, and stored the know-how from these tests in a knowledge base of over 10,000 expert cases. That database is used to help guide new deployments and reduce trial-and-error work.

In practice, this could mean a business spends days instead of weeks tuning a model for performance. In an era where the cost of compute is high and time to market matters, this kind of support can make a difference.

A broader push into AI operations

Huawei’s computing platform announcement was only one of a number of AI-related topics highlighted at MWC this year.

Separately, the company also announced an AI-Native framework for intelligent operations that it plans to release during the event. This framework is intended to help telecom operators and other organisations manage complex networks and AI systems more effectively by using a combination of digital twin simulations and domain models.

Another initiative tied to the event involves software for agent-to-agent communications. On the eve of MWC, Huawei said it would open source a protocol called A2A-T that aims to standardise how autonomous agents in telecom systems communicate. This could help with collaboration between tools and services from different vendors.

These moves fit into a larger theme at MWC this year. Analysts have noted that the 2026 edition of the show has placed heavier emphasis on AI and intelligent connectivity than in past years. The overarching theme of the event has been framed around how artificial intelligence is becoming integral to telecom infrastructure and enterprise services.

What businesses should take from this

It is worth noting that Huawei’s announcements are part of a broader industry trend. Many companies, from cloud providers to networking vendors, are rolling out tools and services aimed at making AI workloads easier to deploy and operate. Huawei’s offering may appeal to enterprises that want an end-to-end service rather than stitching together separate solutions from multiple vendors.

At the same time, the approach still requires significant investment in on-premises infrastructure. While cloud services from major hyperscalers can also provide scalable computing for AI, some businesses prefer to keep data and workloads within their own data centres for performance or regulatory reasons.

Huawei’s platform may help close the gap between large cloud providers and on-premises systems by giving enterprises more support throughout the entire process, from planning and build to operation and model tuning. The real test will come as customers deploy the solution and share their results.

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.