Welcome, Please Sign In

Get in touch with your rep, view past orders, save configurations and more. Don't have an account? Create one in seconds below.

login

Cachengo Edge AI Systems

Take advantage of data outside the data center

AI at the Edge with Cachengo Solutions

With Cachengo, you don’t need a data center for AI and analytics. Improve your operations, gain deeper insights, and deliver more responsive capabilities with Cachengo’s technology for AI at the edge.

Silicon Mechanics has the expertise to help you implement Cachengo solutions that work for your infrastructure.

Cachengo eliminates the need to send data back to the data center or cloud for processing – you get the tools to process data at the source. Get the advantages of faster speed and lower latency, quicker decision making, and enhanced security.

Edge AI Technology

Cachengo’s modular storage and compute clusters can be deployed anywhere, allowing you to deliver AI at the edge with the right combination of security, scale, economics, and performance.

Cachengo offers these advantages:

  • Enhance security with a peer-to-peer network (Hive Connect) that sits on top of Ethernet and is virtually impossible to hack or disintermediate.
  • Increase scale by adding processing power with storage so that capability and capacity expand together.
  • The simplified edge architecture can reduce CAPEX by 5x and OPEX by 4x compared with traditional Intel architectures.
  • Add CPUs, GPUs, and even TPUs to storage to optimize analytics performance at the edge.

Bento Set-Top Box

The Bento box houses 8 Symbiotes. It is not a rack device – it’s designed for deployment in a wide variety of environments.

Features:

  • Low power ARM-based cluster in a box under 50W
  • 48 ARM cores + 32 Mali GPU cores
  • Up to 4096GB of storage
  • Central management Interface for resource, App deployment and orchestration
  • Cloud native and multi-cloud support
  • Support erasure coding and replication data protection
  • Enhanced proprietary SD-WAN solution
  • Just-In-Time Data "Thin-provisioning" data caching support
  • Native ML/AI Capabilities

Pizza 1U Rack Form Factor

The Pizza box is designed as a 1U rack device, providing a high density of 32 Symbiotes. It is ideal for deployment in a standard equipment rack.

Features:

  • Low power ARM-based cluster in a box under 250W
  • 192 ARM cores + 128 Mali GPU cores
  • Up to 16384GB of storage
  • Central management Interface for resource, App deployment and orchestration
  • Cloud native and multi-cloud support
  • Support erasure coding and replication data protection
  • Enhanced proprietary SD-WAN solution
  • Just-In-Time Data "Thin-provisioning" data caching support
  • Native ML/AI Capabilities

Inside Cachengo Technology

Symbiote Independently Addressable Server

The Symbiote drive is an independently addressable server that forms the foundational building block of Cachengo’s hardware. It features an Arm processor with six cores to provide compute capability that meets the minimum power levels often required in many edge environments.

GPUs or TPUs are incorporated in the Symbiote to provide the parallel processing necessary for many analytics workloads, including Computer Vision. Because each Symbiote is an independent device, they allow highly parallel workloads to be spread across multiple nodes. Flash storage delivers high performance/low power data storage and Ethernet networking that enables highly secure peer-to-peer networking.

The Symbiote drives are packaged into server nodes as Bento or Pizza boxes.

Cachengo Portal Software

Cachengo Portal is a cloud-based management software tool that manages the nodes. It is designed as an intuitive, low-touch tool that enables easy provisioning of the Symbiotes and automates many functions to enable efficient management of large-scale deployments.

Management can be a challenge in many large deployments. Cachengo Portal solves that challenge by making it simple to add peer groups and install apps with a simple “point and click” interface.


Expert Included

Our engineers are not only experts in traditional HPC and AI technologies, we also routinely build complex rack-scale solutions with today's newest innovations so that we can design and build the best solution for your unique needs.

Talk to an engineer and see how we can help solve your computing challenges today.