Traditional public clouds encourage lock-in through platform specific tools and large data transfer fees. It is generally not possible to move workloads to different providers without retooling. Since these companies own and operate the infrastructure themselves, even the largest hyperscalers only have so much capacity to go around.
We have invented a new paradigm, where some of the largest public clouds and many others all connect to a our public network. Aggregated together on a level playing field, users may buy compute resources at a significantly lower cost and avoid having their workloads be locked-in to one provider.
This network gives access to an incredibly diverse array of processors - including some of the latest GPUs like the RTX 3090 - at a fraction of the cost of other on-demand options.
The Compute API gives significant control over security such as authentication APIs and encryption. It also makes selecting data residency and compliance certifications straightforward.
The Compute API also gives granular control over how and where your applications are run. Workloads can instantaneously rightsize or switch infrastructure.
The Distributed Computer makes it easy to program for any computing environment, and access thousands of nodes upon request. There is no toolkit like it.
A cloud federation is a union of different cloud platforms that comprises many data centres and heterogeneous infrastructure. Federating these platforms allows users to access greater supply at a lower cost.
The Distributed Computer is a unique cloud federation that makes moving workloads from your on-premises network to the cloud, and even between entirely different cloud platforms quick and painless. It creates the world's first true 'smart grid' for computing resources.
The Distributed Computer is built to Five-Nines standards. If a single node fails, workloads checkpoint their progress and seamlessly move to a new one.
There is an incredible diversity of hardware available through the Distributed Computer, from RTX 3090 GPUs to low-power ARM chips.
The Distributed Computer does not impose any limits on usage or charge a progressively increasing rate, unlike other cloud providers.
CPU and GPU time is sold at or even below-spot prices. The benchmark price for GPUs can be as low as $1,250/year, a major saving over other options.
It is impossible to be locked in to a single cloud or company with the Distributed Computer, as workloads are standardized to be able to transfer anywhere at a moment's notice.
The Distributed Computer meters the use of computing resources with a very simple model, which every provider on the Distributed Computer abides by. Cost is calculated as a factor of compute time and data transfer, without other complex considerations.
Organizations can specify the price they are willing to pay for a particular workload similar to a conventional spot pricing model, or they may choose to pay the clear market rate. Other factors such as enhanced security are transparently factored in to the cost..
You pay only for the actual compute and network resources you consume
The cost of any workload can be predicted with a high degree of confidence, and limits easily applied
The price for the same workload stays within a narrow range, and does not change once begun
As a federation of multiple infrastructure partners, any datacentre or institution is free to join worldwide. Every one of these partners installs technology provided by us to connect to the network and compete to host workloads.
However, the Distributed Computer requires those that wish to host sensitive data to provide additional security documentation. This protects users, who know they will get the same quality of service as they used to from dealing directly with the operator of that hardware.
The Distributed Computer allows users to both buy and sell computing resources through its public network. This makes it possible to sell the compute you have on your own devices back to the network at a profit when you do not need it.
If you already provide public or private cloud resources, or just have a lot of hardware that is going underutilized, joining the public network has many benefits:
The Distributed Computer automatically finds demand for your idle capacity through the public network, safely increasing utilization.
The Distributed Computer's technology allows more operations to take place on the same hardware in a secure fashion than the use of VMs.
By connecting your infrastructure to the public network, you do not need to invest significant amounts in software R&D or marketing to attract customers.