SONM is an infrastructure of a data hub that provides a powerful globally distributed system for general-purpose computing. Attribute to the decentralized open-source platform, SONM boasts variable resources, from data centers to personal computers. With massive computing power and unlimited resources, SONM supports CGI rendering with no software modifications, the current available IaaS product is Rays Render, a rendering service platform based on SONM fog computing infrastructure which outperforms traditional rendering at higher speed and lower cost.
Jumping out of the traditional “pain” zone
Traditional on-demand video rendering can be painstaking. Ever imagine trying to render an animation on a computer? The time span will depend on the length of the video, the complexity of elements like textures, dimensions, shading, effect, etc., and specifications of, for example, hardware, processing cores, bandwidth, and servers. Moreover, complicated details in the frames lead to higher prices, which makes traditional rendering even more discouraging. In the long run, devices and setups would need upgrading to achieve a satisfactory rendering result with higher definition or complexity, let alone the time and energy needed for operating and maintenance.
While SONM’s decentralized system makes things easier. For instance, Rays is able to deliver rendering service with no need for users to own any devices themselves while giving them access to any kind of computing resources at any instant at a low cost. With a drag of a file, the work starts and ends in a surprisingly efficient manner: the task will be allocated to dozens of GPUs at once, so that it won’t be affected by a single node failure, and multitasking will be enabled by over 6000 GPUs; users only need to make requests when needed without having to sit and wait for the response, as the idle computing resources will respond in milliseconds. The platform features a real-time calculator, users will be charged only for the working duration (account for milliseconds). With Rays, users will enjoy a fluid and efficient experience with faster loading time and no waste of resources.
Visions for optimized proximity in local transfer
With SONM’s fog computing backends, Rays naturally benefits from SONM’s outstanding feature: proximity. SONM’s infrastructure distributions are spread around the globe, including PCs, servers, data centers and mining equipment, meaning that unlimited CPUs and GPUs are at hand. Unlike conventional cloud model, where tasks are sent to a centralized cloud before distributing, SONM’s fog allows for decentralized distributions, automatically seeking for the most accessible edge devices to perform certain tasks, easing the data traffic and avoiding possible latencies due to cloud failure, hence accelerating the rendering process. However, Rays won’t stop here. Proximity is where Rays’ essence lies, but also where it makes advances.
Normally, Rays is able to largely narrow the rendering timeframe, while in some rare cases, local internet connections will greatly impact the download speed on remote PCs, especially for extremely large data. In pursuit of better efficacy, Rays is able to optimize proximity with fog computing technology. During a local transfer, SONM nodes will be able to exchange data to decrease the overall loading time, selectively choosing terminals to download resources according to different data sizes. Peer-to-peer content delivery technology would be another solution. Since scene data are the same for all computers that are performing rendering, a single download will serve all the endpoints in the same subnetwork, leading to less congestion and conflicts with other traffic. With all the effort, Rays is able to achieve prominently high speed at minimum cost.
SONM — the backbone of institution-level rendering
On top of offering a smoother and cost-efficient rendering experience for users, SONM sets the goal of building a more applicable system for a wider range of users to reach its full IaaS potential. By harnessing SONM’s web3 infrastructure, users will be able to implement their rendering tasks at a minimal budget, whereas blockchain developers can easily utilize the global web-scale services, to power any metaverse application with higher performance and scalability. In the meantime, SONM will actively seek meaningful partnerships with tier-one counterparts and industry-leading Dapps, to widen business adoption and explore more possibilities in the Web 3.0 era.