Cachengo
Pushing Analytics to the Edge

Today, IT teams within large corporations are leveraging the scalability of cloud-based storage and the ability of machine learning to automate the analytical computation processes and manage their ever-growing datasets. However, in stark contrast to their turn-of-the-century counterparts that stored their data within dedicated, on-premise facilities, businesses in the digital age rely on cloud service providers to process their data and extract actionable insights. And while the reliance on third-party networking partners has its advantages, it does take away the key element of accountability that leaders want to maintain in cases of emergencies such as data breaches, and more. At the same time, most cloud storage ecosystems are not built for data processing, and therefore, have to take their data sets out of storage to run applications and/or analytics. This transfer of data between storage and business applications requires significant resources and expensive IT engineering to perform effectively, which poses a huge roadblock for small-to-medium businesses (SMBs) trying to innovate ML applications using limited competencies.
How can SMBs gain the benefits of intuitive ML-powered processes without having to write complex codes or acquire costly IT engineering solutions, and how can larger businesses take back their data accountability? Tennessee-based Cachengo answers with its Edge AI system that is designed to eliminate the need to transfer data outside its storage for analysis and enable the extraction of deeper insights at the source.

According to Ash Young, CEO of Cachengo, analytics processes today are predominantly dependent on traditional servers that are integrated with GPUs and/or TPUs to boost their ability to parallel process the ever-evolving neural network workloads.

Cachengo approaches neural nets and their complexities from a different perspective through its purpose-built hardware and software offerings that deliver massive parallelism, low power drain, high performance, and tight security at a reasonable price.

Revolutionizing Computational Storage

All facets of the company’s approach are underpinned by its Symbiote drive, which is a revolutionary data management system that is capable of massive highly-parallel processing typical of analytic workloads.Symbiote is a 2.5-inch node with its own dedicated CPU, GPU/TPU, networking, and flash storage capabilities.

“The way we approach data storage and computational need is simple: Why do things the same way, when we can do them in a smarter, greener, and affordable manner?” says Ash Young, CEO of Cachengo.

Symbiote is built to help the company’s clients answer this question in their own unique way by facilitating users to develop and deploy innovative and impactful IoT applications directly from its in-built open source software marketplace. As a result, the system gives its customers the power to truly execute analytics at the source of the data with low latency for optimized edge AI. “With Symbiote, our clients can take control of their data analytics and computational processes by bypassing traditional power, size, or network density constraints,” adds Young.

Using its Symbiote technology, Cachengo has developed its Bento Set-Top Box, which can hold up to eight Symbiotes and deliver the storage and computational cluster of eight traditional data servers in the size of a book. The Bento box also operates at under 50 Watts of power, which combined with its unmatched size can drive significant cost-savings via efficient spatial allocation and lower energy consumption.

Unlocking Innovative ML Applications

Cachengo offers its architecture products in a four-pronged approach, which comprises drives, boxes, networking, and management. The Symbiotes are the company’s drive, which can be assembled into the Bento Box or its Pizza 1U Rack Form Factor product that can hold 32 Symbiotes and deliver computational analytics at ultra-low power. Most SD-WAN technologies simply focus on connecting or extending networks. Concurrently, from a networking standpoint, the company has developed its Cachengo Connect software that enables the linking of the millions of Symbiotes deployed worldwide and allows users to configure them as object storage devices (OSDs) or even as application servers. Cachengo Connect was created with an idea to eliminate the dependency on Secure Shell (SSH) in data communications while facilitating instant messaging and protocol sharing between Symbiotes, no matter where they sit. Lastly, the company provides its Cachengo Portal solution to help users in easily provisioning Symbiotes and seamlessly add new applications from the Cachengo Market, including object storage from MinIO or containers with Kubernetes. “We can even use Cachengo Connect in combination with Cachengo Portal to quickly deploy applications and establish a fully-trackable, secure, auditable data communications between our Symbiotes and boxes,” states Young.