Zenlayer’s hyper-connected cloud powers AI and digital growth in emerging APAC markets
The demand for adaptable infrastructure has never been higher, but as businesses expand their digital footprints, particularly in emerging markets, they face unique challenges and opportunities. Jeff Geiser, VP of customer experience at Zenlayer speaks to EdgeIR about how the company is deploying and optimizing infrastructure to support AI and digital growth in the APAC region specifically.
Zenlayer positions itself as the world’s first “hyper-connected cloud,” a term that Geiser explains is centered around the idea of enhanced network interconnectivity, scalability, and seamless integration. In today’s world, where companies are increasingly blending public cloud deployments with private and hybrid cloud infrastructure, the need for scalable, secure, and fast interconnections is important. Zenlayer has built a network that supports these needs across a global footprint, with over 300 points of presence (POPs) in approximately 110 cities across five continents.
“Our focus has been primarily on emerging markets, with Asia being a key area due to its complex and fragmented infrastructure environment. In these markets, we provide a variety of compute options while ensuring the highest quality network connectivity between these installations, cloud providers, and end users,” Geiser notes.
“We want to focus heavily on providing the highest quality network connectivity between compute installations, cloud providers and to end users. For us, a hyper-connected cloud is a description of being hyper-focused on providing that for customers in emerging markets.”
Additionally, Zenlayer’s port-based product model supports various types of organizations, including managed service providers (MSPs), cloud service providers (CSPs), and enterprises. For MSPs, it extends existing networks to remote locations or public clouds, allowing them to offer services to users in these environments. CSPs benefit by enabling end users to connect to other clouds, facilitating hybrid, inter-cloud, or multi-cloud interconnections. Enterprises can leverage this model to reach a global audience, driving market expansion, brand-building, talent acquisition, and partnerships.
Zenlayer offers a range of networking tools, including a dedicated Software-Defined Network (SDN) fabric, a dedicated IP Transit (IPT) backbone, and service options that are managed through their console.
“What makes our solutions special is the ease with which customers can spin up connectivity on our backbone with just a few clicks,” he explains. “A single port on our platform provides access to all of our network services. This kind of streamlined access is crucial for companies that need to quickly deploy infrastructure and connect various regions seamlessly.”
Navigating the challenges of the Asian market
With over a decade of experience, Zenlayer has established itself in Asia, as Geiser claims, as the region’s number one hyper-connected fabric.
When comparing the Asian market to more mature markets like the US, Geiser points out several key differences. “The US market benefits from being a single country with consistent network connectivity. In contrast, Asia is a vast region with many countries, each with its own providers, regulations, and compliance requirements. This makes it a much more complicated environment to navigate,” he elaborates. “However, this complexity also presents opportunities for companies like Zenlayer to provide value through our expertise and partnerships in the region.”
Zenlayer’s understanding of the local landscape, coupled with their network infrastructure, allows them to help customers overcome these challenges. “We have a decade of experience helping customers succeed in Asia and we have a very rich partner ecosystem (top local carriers, data centers providers etc) there to provide turnkey solutions to our customers,” Geiser adds.
For instance, Zenlayer has partnered with a multinational company that frequently acquires businesses across Asia. This company relies on Zenlayer’s infrastructure to quickly deploy enterprise data services and connect these newly acquired entities to their global backbone, ensuring seamless integration and management of their network.
The role of AI in driving infrastructure demand
Artificial intelligence (AI) is a significant driver of infrastructure demand in emerging markets. Geiser identifies AI as one of the biggest trends shaping the future of digital growth. “AI is driving a need for fabric to connect AI compute clusters in different locations and environments,” Geiser explains. “We see exploding demand for AI data centers, as well as a lot of demand for expertise and know-how in terms of managing the deployment of equipment, the hosting, the connectivity in these dense compute environments for AI training and inference models.
“There are diverse categories of AI inference models that require a highly distributed compute deployment, and low latency connections to end users and devices. That’s where we’re seeing probably the biggest opportunity, and where we’re having companies come to us looking to help them solve some of those challenges. We also see in countries like Indonesia and Saudi Arabia that they each have their own key digital transformation initiatives.”
In response to this demand, Zenlayer has developed tools to support AI deployments. These include managed data center services, guidance on site selection, customs and clearance assistance, and deploying an AI fabric in Asia that supports inter-data center transfer of data for model training.
Looking ahead, Zenlayer is focusing on expanding its capabilities to support the growing needs of AI. Geiser breaks down their strategy into three categories:
Multimodal Learning: Zenlayer is assisting customers in deploying dense compute installations for training diverse AI models. This involves managing everything from hardware procurement to site selection and deployment.
Inference Training: As AI models evolve from training to inference, the need for low latency, region-specific deployments becomes critical. Zenlayer’s network fabric in Asia facilitates the transfer of data between data centers, enabling seamless model training and inference. This is a high-bandwidth, low latency fabric that connects all major data centers in Singapore, Indonesia, Malaysia, Thailand, Japan, the Philippines, Vietnam, etc. The goal, according to the company, is to connect all major compute clusters in Asia.
Edge AI: Zenlayer is investing in edge AI capabilities by making GPUs available on their compute infrastructure. This allows customers to leverage GPUs like the NVIDIA 4090, A100, and H100 for both training and inference workloads, offering a flexible platform for AI development.
“We’ve deployed AI fabric in Asia, where we are doing a very broad deployment of fiber in regions to enable data center transfers of data for model training and then, inference models require low latency connections in a much broader distribution of the model itself, closer to end users and closer to devices,” Geiser explains.
“We want to provide computational power for not only training workloads or distributed training workloads, but where inference models might require GPUs. We want to help provide a platform where we can manage that on behalf of our customers.”
Comments