Now in public beta, Salad Container Engine offers access to thousands of hardware nodes and affordable GPU resources through a convenient deployment portal.
SALT LAKE CITY, Jan. 19, 2023 (GLOBE NEWSWIRE) -- Salad Technologies, a distributed cloud-computing startup based in Salt Lake City, has launched a fully managed container orchestration platform powered by tens of thousands of gaming PCs from around the world. In a rapidly growing market stifled by expensive cloud services, Salad Container Engine is a more affordable, plug-and-play alternative to serverless container products from hyperscale providers.
"Because rising costs are an existential concern for organizations running in the cloud, our platform was purpose-built for seamless integration with existing multicloud configurations, to make container deployments as cost-efficient as possible, with no additional switching cost," said CEO Bob Miles.
According to a 2021 containerization report by DataDog, 50% of organizations have already adopted fully managed serverless technologies into their workflows. With industry adoption of containerized workloads expected to continue at a rapid pace, and SMEs poised to shift 50% of tech spending to cloud resources in 2023, there's real opportunity for alternative providers.
"Salad Container Engine was built to help developers access the resources they need, exactly when they need them, with no DevOps management required," said Daniel Sarfati, Salad's Head of Product.
Developers using Salad Container Engine can securely deploy stateless container workloads across a global cloud infrastructure network powered by privately owned devices. By sustainably leveraging latent computing resources and data-transfer bandwidth from consumer machines, Salad empowers everyday people to compete directly with incumbent cloud services.
Citing a 2021 study commissioned by Dell, Salad believes its platform will appeal to the majority of organizations deploying containerized batch jobs to the public cloud. The startup is now leveraging its competitive price advantage to onboard customers building CPU-intensive compute applications for use cases such as distributed data analysis and AI inference.
In a recent benchmark study, BERT transformer networks running on Salad Container Engine were shown to generate ten times more inferences per dollar spent than Amazon Web Services.
Salad will soon unveil deployment integrations for privately hosted container images and expand its resources portfolio with dedicated GPU support in order to answer a growing waitlist of customers running generative AI workloads and other computationally intensive applications that require 3-D acceleration by such hardware.
To request access to Salad Container Engine, developers can sign up for an account at portal.salad.com and request whitelisting from Salad's onboarding documentation.
Salad is the world's largest community-powered cloud platform. Private individuals share latent computational resources through Salad's desktop application in exchange for personalized rewards from a community storefront. With integrated support for standard container development workflows and thousands of trustworthy, GPU-equipped machines, Salad Container Engine powers cutting-edge compute applications for far less than hyperscale providers. Salad is building a more sustainable and more equitable Internet where everyone can support innovation, profit from their PCs, and activate underutilized compute capacity as affordable infrastructure.
Optimizing AI Inference Costs on SaladCloud Infrastructure.pdf
This content was issued through the press release distribution service at Newswire.com.