Kubernetes, containerisation, serverless computing are emerging trends in cloud computing: AceCloud's Chhabra

As AI continues to evolve, it is reshaping the cloud industry, driving demand for AI infrastructure and specialised cloud solutions. AceCloud, a company specialising in cloud computing services, is adapting to these changes by focusing on AI/ML infrastructure and expanding GPU offerings to support AI workloads, particularly in markets like India where demand is high. In a conversation with TechCircle, Co-founder and Managing Director Vinay Chhabra discussed AceCloud's approach to simplifying IT for businesses with managed cloud services for storage, analytics, and scalable infrastructure. He also outlined the company's strategies, expansion plans, and collaborations, aimed at staying ahead of key cloud and AI trends. 
Edited Excerpts: 

How is AI evolving, and what role is it playing in transforming the cloud space? How is your company positioned to address these changes? 

The highest growth in cloud computing is expected to come from hosted GPUs and AI infrastructure. According to industry reports, AI workloads are growing at a rate of over 35%, which will significantly drive cloud adoption and expansion. When training large language models (LLMs), substantial AI compute power is required for short periods. For example, training a model may demand a cluster of GPUs for two or three days, depending on the workload size, and then may not be needed again for some time. Using the cloud for these tasks is more efficient than in-house infrastructure, which would require purchasing large amounts of hardware that would remain idle between training sessions. As a result, AI workloads are well-suited for cloud environments. 

How does your company simplify complex IT challenges for enterprises with its cloud offerings? 

Cloud computing simplifies IT management by eliminating the need to handle infrastructure and in-house services. Many enterprises today find it challenging to attract and retain highly qualified IT personnel, as skilled professionals frequently change jobs. Maintaining a large, talented workforce in-house is both difficult and costly. Moving to the cloud reduces this complexity, as the infrastructure and uptime are managed by the cloud provider. Organisations can focus solely on their applications without worrying about underlying systems. For instance, if a company uses multiple databases, managing each database requires specialised expertise to ensure availability. Databases are essential, so delays in handling them can impact critical services. By shifting workloads to the cloud, enterprises can simplify operations and reduce overall costs. 

What are some specific use cases or industries where your solutions have had the most impact? 

For workloads that vary significantly, cloud solutions are ideal. If you need, for example, 10 compute instances now and 100 in an hour, the cloud can meet these demands. Cloud providers like us offer the necessary capacity to support variable workloads, charging only for the actual usage time, down to the hour or even minute. In India, our focus areas include object storage and AI/ML solutions. First, with object storage, over 80% of data is unstructured and can benefit from object storage solutions. India’s object storage market is growing at a rate of 20-25%. While many providers offer storage with archiving options, only a few—including us—provide a mix of standard storage and archival solutions (such as AWS Glacier). We are about to launch a service allowing flexible storage options, helping customers save costs on non-critical, long-term data storage needs. 
Our second area of focus is AI and ML infrastructure. Demand for AI and ML solutions is growing rapidly in India, and we are expanding our GPU offerings to meet this need. India’s market has a particularly strong demand for AI and ML workloads, with a high number of companies entering the space and substantial potential for growth. 

What are your thoughts on your current expansion plans, both within India and on a global scale? 

We currently offer cloud services in India at two locations: Mumbai and Noida. We also have a data centre in Atlanta, USA, as part of our established presence in that market. Our plans include expanding to additional regions, including the Middle East, Singapore, Europe, and possibly Australia, within the next one to two years. 

Can you share any recent partnerships or collaborations that have advanced your AI/ML offerings? 

In the AI and ML space, we have several customers, though it's uncertain if these can yet be classified as significant partnerships. In healthcare, for instance, we've provided services to the Institute of Medical Sciences, along with a few other key customers. I don’t have the full list on hand but can share it with you later. Regarding partnerships, we have collaborated with several OEMs who have enhanced the efficiency of our IT infrastructure and offerings. These partnerships help us provide competitive pricing and improved reliability. In the object storage space, we’ve partnered with a U.S.-based company, enabling us to offer a range of storage options, including standard, glacier, and deep glacier storage. These developments are expected to have a substantial impact going forward. 

What emerging trends do you see in AI, ML, or cloud, and how is your company preparing to stay ahead? 

Emerging trends in cloud computing show a shift from traditional virtual machines to Kubernetes and containerisation for hosting applications. Increasingly, applications are migrating to containers with Kubernetes serving as the orchestration platform. Serverless computing is also gaining traction. With serverless, users only incur charges when workloads are active, which appeals to those with variable workloads, as it avoids costs when systems are idle. However, for consistently high workloads, a serverless model may become more expensive, making it less suitable. 

In the AI/ML space, a shift is occurring toward AI/ML as a service. Rather than purchasing dedicated virtual machines with GPUs for AI/ML tasks, organisations can now access an ecosystem where they can quickly provision the required GPUs and run workloads efficiently as needed. This trend aligns with the containerisation shift in cloud computing, enabling greater flexibility and scalability. 

Additionally, there's a trend of companies moving away from hyperscale cloud providers due to high costs. Some enterprises and SMBs are opting for private clouds on-premises, but maintaining an in-house cloud requires substantial technical expertise and resources. A middle-tier cloud provider can be an alternative, offering a balance between cost and performance for non-critical workloads—typically more affordable than hyperscale providers but more robust than private on-premise clouds. This middle-tier option provides a safer and more economical choice for specific use cases.