Decentralized Computers
Note: Data in this section last updated September 12th, 2022
What are Decentralized Computers?
Web3 may represent the biggest computing challenge in history.
Over the next decade, we will likely see the online world expand in both breadth (connecting to billions of IoT devices) and depth (as we begin to realize the vision of the metaverse – a collection of VR-powered virtual worlds).
As such, vast amounts of processing power will be required to create entire digital economies composed of game engines, simulated environments, video streaming, spatial computing, holographic displays and virtual and augmented reality experiences.
Fortunately, blockchains offer a potential solution to this problem through a concept known as decentralized (or “shared”) computing. Shared computing lets individuals lend the unused computing resources on their personal devices (e.g. processing power, storage space, network bandwidth) to those who need it in exchange for compensation through cryptocurrencies.
Shared computing offers several benefits including:
More Processing Power: By relying on a global network of computers, shared computing platforms hope to aggregate more processing power than any single entity could on its own
Faster Speeds: Preliminary tests have shown that distributed computing networks are 20% faster than centralized ones, and this number may continue to grow as networks scale
Lower Costs: Because distributed computing models can aggregate from a variety of sources, they are often significantly cheaper than centralized suppliers. According to Jules Urbach, the CEO of Render, “when you have really cheap GPU power and you want to render something 5,000 times—and the cost is 1/100th what it was before—that becomes tractable.”
More Security: Because identical copies of your information is scattered and stored across multiple nodes, there is little chance of outages, data loss or distributed denial of service attacks (“DDoS” where malicious actors use bots to overload a site with traffic)
Passive Income: Processors can lend their unused computing power to earn passive income
Avivah Lithan, a vice president at Gartner Research believes that shared computing represents “blockchain at its best — peer-to-peer networks where underutilized servers are traded” and compares it to a “Uber” for computers. One of the interesting projects leveraging this model is known as Render Technologies.
How does Shared Compute Work?
Founded in 2016 by Jules Urbach, Render is building a shared computing network that helps “render” visual effects and 3D graphics.
Rendering is the process used by computers to create 3D visual displays with texture and detail. This process is extremely difficult and computationally intensive, and often requires hours to render a single object or frame. For example, according to Pixar’s Peter Collingridge:
“Pixar has a huge "render farm," which is basically a supercomputer composed of 2000 machines, and 24,000 cores. This makes it one of the 25 largest supercomputers in the world. That said, with all that computing power, it still took two years to render Monster's University.”
Render operates as an automated marketplace –a sort of “AirBnB” for computing power – where:
Users who want to lend their unused computing power sign up for the network as “Node Operators” and install the required software
Content creators submit their job requests along with payment in the platform’s native token, RNDR
The Render network automatically assigns these jobs to the appropriate Node, matching the Creator’s need with the Node’s capabilities
Node operators run Render’s “OctaneRender” protocol to complete the work
Once complete, the escrow is released and Nodes receive compensation in RNDR
The network can be used to create several digital assets including videos, NFTs, high-resolution images, metaverse assets and 3D animations.
At the time of writing, Render’s market cap is roughly $300 million.
Who are the Key Players offering Shared Compute?
Shared computing has made significant progress in the last year. In addition to RNDR, other notable projects include LivePeer, Akash and Golem.
Last updated
