This is a project for the Blockchain Hackathon sponsored by BlockGeeks Labs. After that we incorporated and patented the technology, which was acquired by a Quantum Computing Company called Artiste-QB
I started this project in March 2016 as a Google Labs Project. The main idea of Synapse is create a worldwide distributed supercomputer using the potential computing power that is not being used while we charge our phones or tablets, the purpose of this was to use it for research activities (inspired by scientific experiment based at UC Berkeley “seti@home“) or also sell the massive computer power to companies that need it. On the other hand, users that join to the synapse distributed computer, get money rewards while they charge the battery of their phones. My limitation back in 2016 was that my tech was totally a disaster, in order to run each SynapseGL node on phones, I had to root the phone (only worked on Android of course), run a debian virtual machine inside the device (in order to be able to run blender as a client), loosing almost 40% of computing power (CPU only) obtaining catastropic results, so I decided to shut down the project. Now I want to shift the idea, keeping the use the phones while the user is not using it, but the difference is that I want to exploit WebGL standard to do general computing operations and rewards the users with a token based on how much general compute power they gave to the network.
In order to manage the GPU accelerated computations in a universal way, I proposed to use WebGL as a GPU access API. (WebGL is a browser implementation of OpenGL, GL stands for Graphic Library, but also the G from GPU stands for graphics, and we can use it for Fully Turing computing operations or General Processing Computing (GPGPU), so we hacked/exploit WebGL and we could perform general computation task GPU accelerated)
This is an example of a ThreeJS Viewport that is listening for JS Accelerated payload code that exploits shaders to encoding computational single precision point operations, then create a texture that is the representation of the final computation task, we extract pixel per pixel the texture generated, decode each pixel and parse them to the final computational return.
A FEW OF THE MOST WORLD-CHANGING USES:
SynapseGL can distribute the task of rendering CGI, doing in minutes what would take days, for cheaper than anything else. The first successful test of Synapse was rendering photograms using distributed photograms assignation to each phone.
From stock market predictions to big data analysis, Synapse can speed up business computation while lowering costs.
Whether you’re teaching an AI to beat grandmasters or recognize images or human voice patterns.
Synapse can run the world’s first and largest fully decentralized cryptocurrency mining pools. It can also help with cryptography work like integer factorization.
The world of science can benefit immensely from inexpensive and easy access to a powerful supercomputer. From DNA analysis to protein folding simulations to searching for alien life and on and on.
With the synapse network, we can run countless types of decentralized microservices and dapps, which anyone will be able to develop and share (for a profit, if they wish). These are just the tip of the iceberg in the potential use cases. And more will appear as more technologies come to fruition.
Basically, I want to port Golem Network Project (open source) to mobile platforms, and be able to run Synapse in any smartphone or tablet without need jailbreak or rooting the device.