Advancing AI with Civo's GPUs
Discover Fermyon's journey with Civo - enhancing AI inferencing through A100 GPUs.
About Fermyon
Fermyon, a leader in AI inferencing, relies heavily on high-performance GPUs to drive the capabilities of its innovative products, including Fermyon Cloud and Spin. Their latest offering, Fermyon Serverless AI, is revolutionizing AI application development, leveraging the power of GPUs to minimize coding complexity and maximize performance.
At the heart of Fermyon's success is a commitment to high-performance computing, optimized resource utilization, and a strategic partnership with Civo.
What was the problem?
Fermyon's AI inferencing offerings required robust and reliable GPU resources, particularly the high-performance NVIDIA A100, which was in high-demand and scarce in the market. The challenge was not just about finding these GPUs but securing them at a cost-effective rate while ensuring technical support that would allow Fermyon to optimize their use.
Fermyon needed a cloud provider that could meet their technical requirements and align with their business needs.
Objectives
Fermyon was in search of a cloud provider that could meet the following key objectives:
Access to high-quality GPUs
Essential for their advanced AI inferencing, they needed a provider that could offer top-tier GPUs like the NVIDIA A100.
High performance and efficiency
To achieve the best possible performance for AI applications, maximizing efficiency was a priority.
Cost-effectiveness
It was imperative to partner with a cloud provider that could offer competitive pricing without compromising on quality or performance.
How did Civo help?
Civo's provision of NVIDIA A100 GPUs was the cornerstone of Fermyon's successful deployment of serverless AI applications. "The first thing that drew us to Civo, of course, is that you had the right kind of GPUs that we needed," stated Fermyon CEO Matt Butcher. This access to high-quality GPUs was crucial, given the market's scarcity and Fermyon's growing demands.
The partnership between Civo and Fermyon went beyond the mere provision of hardware. It encompassed a high level of ongoing support and collaboration, which was greatly valued by Fermyon. Matt and the Fermyon team appreciated the continuous dialogue with Civo's engineering team, stating, "We've had many conversations over time with engineers at Civo to make sure that we were using the GPU in the most effective way." This close communication was crucial in ensuring that Fermyon could optimize the use of GPUs for their specific needs. Matt further emphasized the significance of this support by referring to it as “top tier support," which illustrates the value Fermyon placed on the technical and customer support provided by Civo.
The operational benefits of Civo's platform, particularly the use of Civo's Terraform providers, were crucial in streamlining Fermyon's infrastructure deployment, saving considerable time and resources. Matt highlighted Civo's straightforward usability with the overwhelming options presented by other cloud providers like AWS or Azure, highlighting the efficiency and user-friendliness of Civo's services.
This ease of use and efficient tooling facilitated Fermyon's smooth transition and supported their growth and scalability in the competitive tech landscape.
The partnership with Civo has significantly boosted Fermyon's AI inferencing capabilities, primarily through access to high-performance NVIDIA A100 GPUs. This access has been a game-changer, enabling Fermyon to achieve greater efficiency and power in their serverless AI offerings.
Matt Butcher summed up the experience with Civo, reflecting on the ease and effectiveness of the partnership: "It's just worked for us, right? We've just been able to use things as they are... This is familiar. I'm comfortable with this." His words reinforce how Civo's solutions not only met Fermyon's immediate technical needs but also played a pivotal role in their growth trajectory.
“The first thing that drew us to Civo, of course, is that you had the right kind of GPUs that we needed.”
