Compute optimization platform Kinesis Network has launched a "serverless feature" that enables enterprises to run workloads across a multi-cloud environment.
The Kinesis serverless solution means that users will have access to infinitely scalable and always available compute services across multiple cloud platforms.
To achieve this, Kinesis pools and redistributes idle compute capacity, including GPUs, giving its users access to compute at around 90 percent less of the cost compared to other providers.
This is particularly useful for AI workloads in the context of high demand and a shortage of affordable computing power. GPUs are available on-demand, and with simple infrastructure management capabilities.
"The compute access shortage is a critical bottleneck for AI innovation and scientific research," said Hamza AK, Kinesis' chief operating officer. "Our platform solves this growing issue by unlocking a vast, untapped reservoir of global computing power, making it accessible to any enterprise or organization. Our team brings a wealth of experience from AWS, IBM, and Microsoft, among others, and we are excited to scale our platform to meet the growing demand for innovative computing resources."
Kinesis was founded by Baris Saydag, CEO and CTO of the company, COO Hamza AK, and Bina Himani, chief product officer and chief revenue officer. AK has previously held roles at Microsoft and Amazon Web Services (AWS), while Khimani has also worked for AWS and IBM.
Seed funding for Kinesis has been raised from "leading investors" including Brilliant Minds.
An early partner of Kinesis is Rare Compute - a biotech platform. "Rare Compute is focused on prioritizing the fight against rare diseases and fueling groundbreaking discoveries that will reshape our understanding of human biology. Our partnership with Kinesis empowers us to focus on advancing generative AI innovations in human genomics related to rare diseases while removing the bottlenecks related to infrastructure," said Stanley Bishop, head scientist at Rare Compute.
"We are fueling enterprise-wide computing transformations, critical research projects, and the next generation of AI startups," added Saydag. "There is strong and growing interest in our secure Compute-as-a-Service solution. We are in discussions with several leading companies and academic institutions about adopting our technology and are excited about our strong pipeline of partnerships."
Earlier this year, Foundry launched a GPU-based Foundry Cloud Platform based on similar idle compute principles, aiming to overcome the GPU shortage for customers seeking to deploy AI workloads.
