Huawei announced the general availability of three new services on its Huawei Cloud during its conference in Dubai this week.
Machine Learning (ML) Ops is one of the new services. Harrison Li, chief cloud solutions architect for Sub-Saharan Africa, explained that some customers said they were struggling to leverage artificial intelligence (AI) capabilities in the cloud.
He told ITWeb that DevOps teams looking to implement an ML solution would have to inject and train the data, and then leverage the algorithm. If they wanted to do it again, they needed to start from scratch, he added.
“We’ve tried to make sure every deployment [of AI] is in a pipeline. The customer just has to input the data to leverage the data store and the process will then run automatically. It’ll be easier, and they’ll save a lot of labour.
“We’ve tried to analyse a typical process in which ML teams build a model, and then we’ve built a pipeline with all the key points in that process. All you have to do is enter the data, and the process will run automatically.”
Another new service is the serverless offering, which is similar to the Amazon Web Services Lambda computing platform.
The customer can choose a preferred computer language and build their computing functions using the Huawei Cloud.
“You know where the infrastructure is and how big it is,” said Li.
“For example, I need four vCPUs [virtual central processing units] and 8GB memory; it runs, and when it’s finished, it shuts down,” he explained, adding the service is charged by the second.
The third service now in release is Metastudio, which will help customers build online digital content, says Huawei.
”The metaverse is so popular right now, and we understand the processes you need to build this digital content,” said Li.
This solution involves optimising the graphic processing unit to ensure content is rendered in high-definition, and then quickly delivered to an edge device, such as augmented reality/virtual reality glasses, from the cloud.
The company used Metastudio to design what it calls its first “digital employee”, YungShen, also known as Sarah.
She was created using automatic AI modelling, and AI voice modelling and rendering, and was present at the conference, albeit confined to a double-sided touch-screen close to the bar.
Share