Computes is Not Your Daddy’s Blockchain Computer

We are building a global, decentralized and distributed, mesh supercomputer from idle CPUs and GPUs. Sounds like another blockchain project, right? Think again!

Blockchain technology has been around for nearly a decade. We are building a new decentralized and distributed platform on what we believe is the next generation of blockchain technology —a DAG-based distributed ledger and decentralized queuing technology called Lattice.

When we set out to build our new decentralized and distributed supercomputer a couple of years ago, we evaluated various blockchain technologies but encountered several concerns with the current state of these technologies:

• The blockchain is currently too slow for supercomputing. At the time of our evaluations, Ethereum was running at approximately 12 transactions per second.
• The overhead of storing and synchronizing blockchain transactions on edge devices was too high especially for IoT and mobile devices.
• Proof of Work (PoW) computations are unnecessary for B2B supercomputing. Why waste CPU and GPU cycles on PoW computations when they could be used for computations such as cancer research?

Being huge fans of the Interplanetary File System (IPFS) project, we quickly realized that this protocol was far more powerful than simply sharing files in a peer-to-peer (P2P) network. Overnight, IPFS became the core technology stack of our next generation decentralized and distributed platform, Computes!

In addition to IPFS being Computes’ native file system, here are a few other IPFS features used in the Computes architecture:

• IPFS has the ability to create public and private mesh networks. We use this feature for creating private mesh computer networks behind large organizations’ firewalls.
• IPFS has a build-in DAG for linking content via hashes. We use the DAG for linking tasks (transactions) queues in Lattice and managing input/output/log files.
• IPLD, IPFS’ Linked Data model, is used to create a CRDT (conflict-free replicated data type) database system. This is used to keep track of hash pointers to current DAG entries and files.
• IPFS’ PubSub is used extensively within Lattice to communicate state of queues and computations across all of the nodes on the mesh computer. PubSub is also used for inter-node communications allowing nodes to interact with one another organically in a P2P fashion on more complex computations or coordination of efforts.

As a result of using IPFS as our core technology stack, our Computes nanocore technology is extremely small and fast and incurs no wasted computing costs associated with proof-of-work or proof-of-stake mining operations. We can “activate” our mesh computer within corporate networks on demand with no additional infrastructure or tokens.

Other benefits of mesh computing (over blockchain computing) include organic swarming networks with P2P discovery and P2P communications. This benefit allows our Lattice queues to find the right computing resources with the right MIPS, right availability, and right location to compute with little to no latency and be closest to the data streams to achieve real-time AI/machine learning as well as massively parallel software-defined supercomputing when and where it’s needed.

Computes nodes organically connect, communicate, and compute. There is no additional infrastructure required to deploy computes’ nanocores in your organization or job site. Computes can even run on existing IoT devices connected to edge networks. Since Computes is decentralized, it’s fault-tolerant computing by default. (In fact, there is no way to stop Computes from computing.)

Contact us (hello@computes.com) for more details. Stay tuned to this blog for more exciting information about our new technology stack and development progress! You can also reach us on Twitter, Facebook, and GitHub.