A Distributed Supercomputer
DCP is the next-generation supercomputer.
DCP is a web-based protocol for computational resources. It is powered by organizations and individuals with idle compute power, who earn rewards and reduce costs.
There is a growing demand for machine learning and other highly parallelizable problems.
DCP enables JSON and Javascript-based computational packages to be sent and received using web technology.
Anyone in Canada will be able to connect any device with an internet browser to perform computations and earn Distributed Compute Credits (DCC's). DCC's will then be exchanged for currency.
The Technology
Distributed Compute Protocol
KDS is creating the Distributed Compute Protocol which enables JSON and Javascript-based computational packages to be sent and received on a distributed network using web technology.
DCP Math
We are making advanced mathematical libraries, including complex arguments and special functions available to javascript.
DCP Physics
The web currently lacks a comprehensive physics engine. We will enable physics-based simulations from the browser.
DCP Shaders
GPU computing is ideal for a wide range of floating point math and physics simulation scenarios. KDS aims to provide an array of readily available shaders for calculation and rendering.
Emscripten
Emscripten is a C++ to Javascript compiler using asm.js; a performance based subset of javascript. Unity engine uses Emscripten to export their applications to the web. It has been used to port Unreal Engine 3 to client side browsers.
WebGL 2.0
In early 2017, Khronos group released WebGL 2.0, bringing significant improvments to browser graphics. "Incorporating Compute Shaders into WebGL is the working group's top priority." ~Kenneth Russel
Javascript
Javascript is the most popular and fastest growing programming language. It's standardized and secure, making it ideal for distributed applications.
Extensions to Support
Industry Standards
MatLab & Maple
Widely used by students and researchers all over the world, MatLab extensions will allow easy access for math students without network experience.
COMSOL
COMSOL is the industry leader in multiphysics modeling and simulation software. Users of this power-hungry software will benefit from our planned networking module.
Blender
We want to enable smaller users and projects to access time-saving power. Parsing blend files for use with WebGL will enable model pre-processing and live rendering for web applications.
Alpha release
The Berkeley Open Infrastructure for Network Computing (BOINC) is an open source volunteer computing platform. It is primarily used for research, including famous projects such as SETI, LHC and Einstein@home.
BOINC has been running since 2002, boasts dozens of projects, hundreds of thousands of miners, and over one million connected machines. At 16.5 PFlopS, they are the 7th fastest supercomputer on earth. *comparing BOINC's April llth, 2018 24-hour average with the top500 list November 2017
While BOINC has a credit system, those credits hold no economic value. Our alpha, called SPARC uses BOINC's open source API to monitor credits distributed among a whitelist of projects, and reward users who sign up to our system with tokens.
Live alpha (SPARC)
Usecase Example - Applied Electrodynamic Simulation
Research by Daniel Desjardins, PhD
Dan represents an example of a small researcher working with a limited team who does not have access to supercomputer time granted by governments and universities. There are tens of thousands of such researchers across the world who would benefit from easy and open access to computational resources.
At the Royal Military College of Canada (Kingston, Ontario), Dan conducts research in applied electromagnetics. He develops solutions describing the complicated interactions arising between magnetic fields and nearby conducting objects.
Dan has recently derived exact mathematical solutions that describe induced current densities and magnetic levitation forces produced by an electromagnetic coil carrying an alternating current and traveling along the inside a conducting tube. His research is leading to innovative electromagnetic propulsion technologies, similar to the research that underpins revolutionary ideas like Hyperloop One.
To view his results, Dan needed to evaluate his solutions numerically over a grid of 4221 points, which took 5 hours to compute. That computational time increases exponentially when considering additional variable parameters such as coil dimensions, tube thickness and conductivity, coil velocity and current frequency.
Dan would like to be able to efficiently compute these solutions for various coil configurations, tube properties, and other parameters in order to optimize propulsion effects, but his computer is not powerful enough. The more computational power he is provided, the higher precision and greater variety of simulations he can perform. More avenues for exploration result in a greater need for computational resources. This demonstrates how supply creates demand.
The induced current density is given by:
$$\textbf{j} \left( r,z,t,v \right) =-{\frac {\mu_{0}\,\sigma\,vni}{R_{i}} \int_{0}^{\infty }\!\Phi \left( k \right) \left( \cos \left( k \left( z-vtz \right) \right) \Im \left( {\frac {A{{K}_{1}\left( \Lambda\,r\right)}+B{{I}_{1}\left(\Lambda\,r\right)}}{AC-BD}} \right) -\sin \left( k \left( z-vt \right) \right) \Re \left( { \frac {A{{K}_{1}\left(\Lambda\,r\right)}+B{{I}_{1}\left( \Lambda\,r\right)}}{AC-BD}} \right) \right) \,{\rm d}k}$$
with definitions:
$$\Phi \left( k \right) =\frac{\sin \left( {\it kl}/2 \right) }{k} \left( b \left( {{I}_{1}\left(kb\right)}{{\rm \bf L}_{0}\left(kb \right)}-{{I}_{0}\left(kb\right)}{{\rm \bf L}_{1}\left(kb\right)} \right) -a \left( {{I}_{1}\left(ka\right)}{{\rm \bf L}_{0}\left(k a\right)}-{{I}_{0}\left(ka\right)}{{\rm \bf L}_{1}\left(ka\right)} \right) \right)$$ $$A=\Lambda\,{{I}_{0}\left(\Lambda\,R_{o}\right)}{{K}_{1}\left(k R_{o}\right)}+k{{K}_{0}\left(kR_{o}\right)}{{I}_{1}\left( \Lambda\,R_{o}\right)}$$ $$B=\Lambda\,{{K}_{0}\left(\Lambda\,R_{o}\right)}{{K}_{1}\left(k R_{o}\right)}-k{{K}_{0}\left(kR_{o}\right)}{{K}_{1}\left( \Lambda\,R_{o}\right)}$$ $$C=\Lambda\,{{K}_{0}\left(\Lambda\,R_{i}\right)}{{I}_{1}\left(k R_{i}\right)}+k{{I}_{0}\left(kR_{i}\right)}{{K}_{1}\left( \Lambda\,R_{i}\right)}$$ $$D=\Lambda\,{{I}_{0}\left(\Lambda\,R_{i}\right)}{{I}_{1}\left(k R_{i}\right)}-k{{I}_{0}\left(kR_{i}\right)}{{I}_{1}\left( \Lambda\,R_{i}\right)}$$ $$\Lambda=\sqrt {{k}^{2}+Ik\mu_{0}\,\sigma v}$$
Behind The Scenes
We're Hiring
Specialist in Computational Mathematics (Kingston, ON, Canada)
We are always looking for talented developers. If you think KDS is right for you send your pitch to jobs@kingsds.network
Daniel Desjardins, PHD
CEO
An experienced communicator and a natural team-leader, Dan is a determined, highly creative and innovative scientist-entrepreneur. He obtained a PhD in Physics, specialization in computational electromagnetics, from Queen's University. Dan has 14 years of professional military experience, having served as a CC130H Hercules military search and rescue pilot, Unit Public Affairs Officer administering media outreach and inquiries, and Assistant Professor with the Department of Physics and Space Science at the Royal Military College of Canada. Dan fulfills his passion for team-building and scientific innovation at KDS.
Greg Agnew
Founder, CTO
Greg has always been at the forefront of technology, having mastered Ethereum and full stack development. Spending his spare time building chess engine style games in web applications, he is an expert in pushing JS to the limits.
David Clark-Joseph
CFO
David is an experienced executive in private equity and portfolio company operations (CFO), with expertise in client negotiations, deal and entity structuring, underwriting and due diligence. He has served in Fortune 500 corporate finance (FP&A) and corporate development, with a public and private firm track record of leading organic and inorganic growth, acquisition selection, modeling, due diligence, integration, synergy realization and performance measurement.
Andrew Pohran
COO
Andrew is an experienced project manager, focused on business operations with a passion for data-driven decision making. He has experience in both the private and public sectors, having worked in the Managed IT Services channel and served as an air combat systems officer in the Royal Canadian Air Force. Andrew manages strategic and long-term partnerships, business processes, and performance objectives.
Wes Garland
DCP core developer
Wes is a software developer, project manager, product manager and team leader with more than 20 years of experience in the industry. He has successfully delivered products in the telecom, gaming, real estate, publishing, mobile and product training verticals. With a proven track record in secure and reliable software design, Wes has delivered products used by national cellular and messaging carriers in five-nines uptime environments. He has been interested in server-side JavaScript platform development his entire career, and served as a member of the Securable Modules working group which eventually became CommonJS. He is the primary author of GPSEE, a Mozilla-based JavaScript development environment (a precursor to NodeJS), BravoJS (a CommonJS Modules/2.0 extensible module system), and the CommonJS Modules/2.0-8 draft specification. Having developed and lead teams using these products since 2007, Wes brings significant expertise in key areas to Kings Distributed Systems.
Matthew Palma
DCP core developer
Matthew is a Computer Science graduate from Queen's University (Kingston, Ontario) who is passionate about software development, blockchain, and the exciting future of the decentralized world. Always looking for ways to improve himself and his work, he has been valuable addition to the team. In his free time he enjoys developing games, and participates in the occasional game jam.
Conrad Bielicki
DCP core developer
Conrad is a software developer specializing in UI development. He has a background in educational video and website production. He has worked in Kingston on projects for the Perimeter Institute, the Canadian Museum of Nature and Show Communications.
KC Erb, MS
Math library and application developer
KC is a physics PhD student at Charles University (Prague, Czech Republic). His research has included computationally intensive fields such as medical image reconstruction and crystallographic structure determination. He has also worked professionally in a variety of capacities as a software engineer throughout his postgraduate studies. He is passionate about improving the quality of software in science.
Mohamedd Salah Hussein
Math library and application developer
Dr. Mohamed is post-doctoral fellow at the Royal Military College of Canada (Kingston, Ontario) specializing in applied research, radiation physics, and information technology. He develops mathematical models and compute algorithms for use in numerical problems, boundary value, and integral equations. Specifically, his expertise lies in high performance parallel computing to simulate radiation transport problems. “I am very excited to be contributing to this project because this IS the future of scientific computing"
Geoff Klein
Math library and application developer
Geoff is a PhD Candidate at University of Toronto with an extensive knowledge of electromagnetic theory, nuclear physics, and nuclear reactor theory. His vast experience with Finite Element Method (FEM) modelling (COMSOL Multiphysics), numerical and analytical techniques, and parallelization make him a valuable member of the Math Team, developing tools for this next generation compute platform.
Christopher Colley, PhD
Math library and application developer
Christopher holds a PhD in Aerospace Engineering, the focus of his research having been the development of numerical methods for applications in scientific and engineering computing. His strong background in computational mathematics and expertise in Python and JavaScript make him a strong addition to the KDS development team.
Jason Kinnear
Venture4th Communications
Jason Kinnear has over 17 years of experience developing and executing results-oriented Canadian, North American and international marketing communications campaigns on behalf of some of industry's best-known technology companies, including Oracle, Adobe, Hewlett Packard, Citrix, TELUS, Symantec and Intuit. Jason spent over eight years as SVP and National Technology Sector Lead for one of Canada's largest communications firm. During this time, he built and grew the practice to what is now one of the top Technology Practices in the country. Jason is also a strong advocate and supporter of Canada's startup ecosystem. He has worked with Communitech and is a frequent advisor to entrepreneurs at MaRS. Jason holds a Degree in History from the University of Western Ontario, a Diploma in Public Relations and Corporate Communications from Seneca College, and has taken a variety of television production workshops through Ryerson University's school of broadcast journalism.
Andrew McColl
Business Development & Partnerships Consultant at OKLink, Hong Kong.
A Commerce student at Queen’s University, Andrew brings expertise in international business development and strategic partnerships to the KDS team. At OKCoin in Hong Kong, he executed innovative ideas on a global scale with a diverse range of partners. Andrew’s experience and connections make him ideal for driving adoption of the DCP network both in Canada and abroad. An avid technology enthusiast with a passion for Blockchain and distributed systems, Andrew is also a valuable liaison with the student community at Queen’s and other top Canadian universities, facilitating the transfer of talent, ideas, and publicity to KDS.
Jack Troughton
Systems Administrator
Jack Troughton has twenty years experience working in networking and security, including nearly a decade in tier 1 data centres maintaining a five-nine uptime telco mail-to-text system with over ten million users. In his spare time he works in professional audio and music.
Aaron Grinhaus, LL.B., J.D., LL.M. (Tax)
Grinhaus Law Firm
Aaron is an experienced business, tax and Fintech consultant, who advises clients on Fintech strategies, including the use of blockchain technology, whitepaper/ICO/ITO regulations and structures, smart contracts and cryptocurrency to reduce business costs, hedge institutional friction and expedite capital raises and cross-border wealth transfer transactions.
Dan Wynngaarden
Accountant
Ron Kochman
Managing Partner, Doyle Capital Management
Involved in angel investing and advisory services to startups in the cryptoasset space. Ron represents a number of high-profile investors with specific interests in digital infrastructure projects such as those being developed by KDS.
Kristine Spekkens, PhD
Associate Professor at Queen's University
Kristine Spekkens is an associate professor in the Department of Physics and Space Science at the Royal Military College of Canada (Kingston, Ontario), and cross-appointed in the Department of Physics, Engineering Physics and Astronomy at Queen's University (Kingston, Ontario). Her research in astronomy focuses on understanding of the structure and evolution of nearby galaxies in a cosmological context. She is particularly excited at the prospect of using future radio telescopes to map the distribution and kinematics of atomic hydrogen in galaxies in diverse cosmic environments, and the computational challenges associated with these ambitious surveys.
Claudia Nakhla, CFA
Director, Business Strategy Development at TELUS
Claudia is a strategic thinker with a passion for applying rigour and strong analysis to planning and idea evaluation, turning the big picture into practical, real initiatives. At TELUS, Claudia manages the Strategy & Planning team in its Consumer & Small Business segment, working with stakeholders to develop overall strategic direction, scorecards, special projects, and reporting of special topics to the Board. Claudia has also managed strategy, business case development, partnership identification, and competitive intelligence for its wholesale and enterprise businesses. She has a keen interest in the dynamics of the global and Canadian telecom markets and innovative ideas that will change the landscape for TELUS.
Gunnar Blohm
Academic Community Advisor
Gunnar is an associate professor for Computational Neuroscience at Queen's University (Kingston, Ontario, Canada). His labtakes a computational approach to sensorimotor neuroscience. Computational modeling is essential for our understanding of neural mechanisms because they formalize hypotheses and make new testable predictions to conduct exciting research lines. "I am very excited about this project because its visionary approach promises to transform data analytics, computational capacity, and accessibility of computing power with tremendous benefits for science and society."
How to Connect
Office & Snail Mail
Kings Distributed Systems
303 Bagot St, Suite 403
Kingston, Ontario, Canada
K7K 5W7