Every day, humans create 2.5 quintillion bytes of data — that is 25 followed by 17 zeros. This data comes from all the activity we do on the internet: be it using social media, gathering data from satellites, making a financial transaction, or using a GPS signal. This amount of data can be as high 10 million blu-ray discs, the total height of which would be equal to four times the high as the Eiffel Tower.
But how much of data are we actually using? The staggering numbers clearly suggest what we require is much less than what we generate. But even then, the data distribution is heavily uneven. For instance, scientists might find it difficult to obtain large amounts of computational data for his/her research, while one in need of only a few megabytes might end up getting a gigabyte worth of computational power.
Cloud Computing, to some extent, has tried to solve this issue. But in the end, it has fallen prey to increased monopolization, poor data security and privacy, and constant downtime. It is the entire concentration of data in one place which makes Cloud Computing an uneasy choice for scientific and neural network projects, site hosting, and other applications. It is purely centralized!
This will likely change with the arrival of Fog Computing. It is a decentralized infrastructure in which computational resources are distributed among the small data centers. In simple words, Fog Computing enables people to use smart devices such as smartphones, laptops, IoT-enabled machines as computing resource hubs.
Key Advantages of Fog Computing
Fog Computing supports high performance applications and focus on real-time interactions than batch processing. The devices can communicate with each other and make decisions independently from clouds. Specifically for the Internet of Things, it is clear that important data, computing power, memory and applications are much more effective than ever before.
The most important advantages of Fog Computing are:
System-wide security: secure edge devices protect multiple access points and detect errors;
Low network latency: Reduction of data transfers by processing directly at the collection points;
Reduction of data transmission: data are filtered/cleaned before they are sent to central processing systems;
Resistance to spread and heterogeneity: multiple fault points make the system robust against redundancy;
Context Awareness: Data is processed where the correct context to the data, in relation to location, time, etc., exists.
Many major vendors, including Cisco, Dell, and HP, are pushing their plans to patent Fog-based Edge Computing models. While these organizations are highly centralized, and are likely to bring users into the same loop of business monopoly, there are efforts in the decentralized industries that are more focused of making users the actual contributor.
SONM Project
SONM, which stands for Supercomputer Organized by Network Mining, is a decentralized worldwide fog supercomputer for everything from general purpose computing to CGI rendering and scientific calculations. The Ethereum project takes a leap from centralized cloud service models, and proposes to build and implement a fog computing structure — a decentralized pool of devices, all of which are connected to the internet (Internet of Thing/Internet of Everything).
At the same time, SONM is taking steps towards replacing Proof-of-Work cryptocurrency mining, which is evidently controlled by miners with expensive ASIC machines.
Miners and Buyers
SONM system is all about using mathematics as a means to connect people looking to get computing power and hubs that are offering it. Miners, or computational power owners, can lend a portion of their device’s power for some useful calculations and processing real tasks. SONM utilizes fog computing structure to bring solo mining back to its prime.
There are lots of miners with GPU mining farms becoming useless due to the increased Proof-of-work mining difficulty (even for altcoins). In recent years, being a part of a mining pool has been the only way to guarantee profit from mining. But even in doing so, this profit is so small that sometimes it doesn’t even cover the cost of electricity spent for PoW mining.
With SONM, miners get the opportunity to make returns by simply serving calculations to the network. The project puts a special focus in the elimination of all the confusion there is about migration from PoW to Ethereum’s PoS — each miner is suggested most profitable applications and tasks for his hardware. Miners can use anything from a smartphone to a supercomputer for SONM fog computing. All they need is to set up a mining client application and run it.
Buyers, on the other hand, get to receive a cheaper and more secure alternative to traditional cloud computing services (AWS, Microsoft, Google Cloud, etc.). SONM has a hybrid architecture, and therefore supports any kind of computational tasks without facing Ethereum’s “out of gas” problem.
Think of a scenario in which a low-profit organization such as a public library, which has much unused computing power, is at one end, and researchers in need of computing power are at another end. SONM can connect them both; which means, extra income for the public library and task fulfillment for the researchers.
Are you still thinking how can you use your unused computing power? Just remember, you are competition to those big and clumsy corporations in yourself.