Kings. Distributed Systems

Transcription

1 Kings Distributed Systems Page 1

2 Table of Contents Overview 3 The Problem 3 The Solution 4 Proof of Market 4 Science Power and Research Coin 4 Variable Hardware Alpha 5 BOINC 5 Miners 6 The Next Step Release Candidate 7 Proprietary Distributed Network 7 Patents 7 Flexible for Miners 8 Application Modes Ethereum Token Factory 9 Ethereum 9 SPARC Token Factory 10 Crowdsale 10 Initial Coin Offering 10 Budget 11 Expectations 11 Revenue and Growth 12 Competitors 12 Howey Disclaimer 13 Team 14 Glossary 15 External Links Page 2

3 Overview Kings Distributed Systems believes currencies need to solve problems to retain their long-term value and stability. The Problem KDS has identified inefficiencies in blockchain currencies. In the current ecosystem miners are using computing power to hash blocks without regard for the results of their computations. They compete to hash transactions looking for a number below the set global difficulty. The first miner to prove his number is below the difficulty is rewarded with a predetermined amount of coins while the rest of the community begins to work on the next set of transactions. 99% WASTE As a result of this process all hashes computed above the difficulty, or by the rest of the ecosystem are discarded. The amount of work discarded by Bitcoin alone amounts to 5,000 peta-hashes; equivalent to 64,000x the combined power of the Top 500 Supercomputers. Networks such as Bitcoin are throwing away power instead of selling to industry or donating to research. The Solution Supercomputing power is a growing industry with immense demand and we will tap into this activity rewarding our miners. Instead of hashing to determine proof of work we are rewarding real work distributed in the form of work units. 99% These work units are mathematical or rendering packages that are being created by research or commercial projects in need of power. By running a network which allows miners to download packages and upload them automatically we can streamline and track their work and reward them with coins representing the value generated by that work. RETAINED Page 3

4 Proof of Market The alpha version of our network connects to the Berkeley Open Infrastructure for Network Computing (BOINC). Since 2002 the University of California Berkeley has been distributing research work across a volunteer computing platform. Over these 15 years they have helped numerous projects get access to valuable processing resources. In its present state they boast over two hundred thousand volunteer miners running across upwards of one million machines. 1,128,897 MACHINES At 17.1 PFLOPS they are the 3rd fastest supercomputer on earth. Our alpha is designed to reward these already working miners with a coin to represent their completed work. The release version of our product will run a proprietary network and open up these distributed applications to commercial buyers. We will continue to support the research community by donating a portion of the network power to their projects in the form of coin grants. These grants are deposited directly into their network accounts and made available to miners each week. This provides a stable base of work for the economy. Science Power AND Research Coin (SPARC) Science Power And Research Coins or SPARC (read: spark) are the token upon which our network runs. Projects who need power purchase SPARC from an exchange and attach them to their work. Miners compete each week for available tokens for each project. At the end of the week the tokens are divided among the miners relative to their work completed. These tokens can then be exchanged directly for computing power from the network or traded on an exchange. Variable Hardware Instead of Variable Pricing Whereas Amazon EC2, Goolge Cloud Compute and other services vary the price depending on their system loads, our network varies the hardware dedicated to your service. Watching for low supply and spinning up new instances is difficult, and your funding is static. We allow you to set your price point and let the network compete for available jobs. This ensures 100% network usage, and potentially high rates of power at low cost without added setup for your project. Page 4

5 Alpha The alpha version of our product is slated for release in September The release candidate is scheduled for Q1 2018, however we will continue to run Alpha on the BOINC network past this date. Berkeley Open Infrastructure for Network Computing BOINC started as a client for SETI@Home and has grown to supply power to dozens of different projects. They supply work units to their miners in the form of executable packages created by the individual projects. Work units come in various forms focused at different machines. The majority of the projects run GPU intensive floating point calculations while a smaller portion are best suited for CPU power. KDS connects to the WebRPC API exposed by the BOINC network and rewards work units completed with Science Power And Research Coins. These SPARC are tokens which exist within the Ethereum network and can be traded on exchanges or held for our release version. They will be accepted by KDS for project power from our network. BOINC is an open source application under development, and we hope to support them by attracting miners through our alpha release, and on our release candidate by donating SPARC to their projects. Miners The miners are volunteers from across the globe who wish to use their extra computing resources to help push humanity forward. They can use almost any computer to connect to the BOINC network and complete work. There are currently just under two hundred thousand active miners on the BOINC network contributing over 17 PFLOPS. Since ,521 MINERS Page 5

6 KDS IS The Next STEP Gridcoin is a cryptocurrency that has been operating on the BOINC network in the same way we plan to run our Alpha. Currently Gridcoin is limited to miners who belong to their project group, whereas SPARC will be available to all miners regardless of affiliation. They have existed since 2015 and as of June 26th have a market cap of 25 Million USD, with a daily trading volume of 700 thousand USD. $25 MILLION USD MARKET CAP While impressive for a hobby coin, it is a low valuation compared to the available market. Gridcoin is a fork of the Peercoin source which is itself a fork of Bitcoin. Since they have no salaried developers their development is slow, and their chain technology is outdated and as such we plan to launch as a more competitive coin. $448,613 DAILY VOLUME The Alpha release will run in parallel to Gridcoin, and miners on the BOINC network will be able to collect both Gridcoin and SPARC therefore we do not consider them to be a direct competitor. We hope to attract these miners which are motivated for coins to our release network by rewarding them with SPARC during our Alpha. Page 6

7 Release Candidate The release candidate is scheduled for Q Proprietary Distributed Network The KDS proprietary network is a peer-to-peer application that connects miners with projects. Projects will purchase SPARC from the exchange and apply them towards their work to attract research power. The application accepts a package of files from projects including a data Blob, CPU Service, and/or GPU Service. The GPU Service consists of one or more OpenGL and/or OpenCL Shaders which can be configured to calculate floating point math, or render scenes. The application wraps these services to protect the miners and exposes a limited API for communication between services. The Blob is split into chunks and fed into the service upon request allowing for a time-out function to guard against stalled services. Completed work is written to a Results Blob. When the work is complete a hash of the Results Blob is sent to the KingsDS Network to be confirmed with other miners. After verification miners will upload their results directly to the project's server. There are four modes which the project can be configured to allow miners to select optimal projects for their machines. Patents We are in the process of filing and expect to receive patents for the technology involved in the network and application. We are expecting to file these patents prior to our ICO and will update this section when patents are pending. Flexible for Miners There are four application modes which help projects select the best resources for their needs. Categorizing work units into various types allows miners to prioritize projects suited to their machines. On the next page we outline the various modes. Page 7

8 Mode 1 DATA CPU GPU CPU RESULT The CPU Service accepts chunks from the Blob, which can than be configured for the GPU Service. Once sent to the GPU Service and work is completed the results are passed back to the CPU Service for interpretation. The CPU Service is responsible for writing results to the Result Blob. Mode 2 DATA CPU GPU RESULT The CPU Service accepts chunks from the Blob, which can than be configured for the GPU Service. Once sent to the GPU Service and work is completed the results will be written automatically to the Results Blob. Mode 3 DATA CPU RESULT The CPU Service accepts chunks from the Blob and completes the necessary calculations. The CPU Service is responsible for writing results to the Result Blob. Mode 4 DATA GPU RESULT The application feeds chunks of Blob data directly to the GPU Service which than writes automatically to the Results Blob. Page 8

9 ETHEREUM TOKEN FACTORY Ethereum Ethereum is a decentralized blockchain platform. They allow smart contracts to execute across the network for any reason. One of these purposes is to allow ledger based tokens to create, transfer value. We will be using the Ethereum network to create and distribute SPARC. Token Factory The SPARC factory is responsible for distribution of tokens across the working miners. Each project may assign a number of SPARC per week to their project. Miner The miners will select one or many projects which they find most valuable to complete work for. At the end of the week the total work completed by each miner for each project will be tallied. The SPARC are then distributed proportionally to miners for the work they have done relative to the total work completed on that project for that week. Miners may select multiple projects and give priorities to their selections to maximize their utility. This is an auction format that will balance the supply of miners against projects demand for power. KDS Server Verifies Work In order to provide a baseline pricing for the auction a number of SPARC grants will be given each week to a whitelist of projects doing research for science. These grants will be factory generated tokens thereby increasing the circulation of tokens. This is our inflationary mechanic to maintain a healthy supply in the ecosystem. 15% of coins used in the auction are returned to the foundation to help support and grow the network. The grants will be static amounts totaling 3,000,000 SPARC per week. We expect this to create 10% growth in circulation during the first year, naturally shrinking each year thereafter. Page 9

10 Crowdsale The crowdsale begins on the morning of October 1st 2017 at 11AM EST. Initial Coin Offering (ICO) KingsDS is raising funds for the continued development and scaling of our network. In return for Ethereum, buyers will receive an initial supply of SPARC that can be traded on the exchange or held until the release candidate for use on the network. Funding Goal 1 $5 Million USD Funding Goal 2 $25 Million USD Funding Goal 3 $50 Million USD Ether : SPARC Ratio 1 : 10,000 Approx SPARC Price $0.035 USD ICO Max SPARC Supply 2,500,000,000 Tokens are long-term assets and 5% of SPARC sold as such we need to ensure a Developer Equity minimum operating budget for at least 5 years. We have budgeted a minimum $1 Million USD per year for this purpose. We expect to raise more based on enthusiasm for similar projects and have set funding goals at $25 and $50 Million. The funds raised will go towards development, scaling, and attracting business for our miners. The approximate SPARC price is at time of writing. The SPARC to Ether ratio is our static measure, and as such the max SPARC supply may change at time of ICO to represent our USD goals. Budget Since the funding is variable, we have created a budget based on our three goals. These numbers represent a rough figure as a percentage of the ICO. We expect in some cases to spend less then these amounts, and have budgeted conservatively to ensure the continuation of the foundation. ICO Funding Goal $5 Million $25 Million $50 Million Release Candidate 1 year 8 months 6 months Developers 5 Developers 10 Developers 15 Developers Administration 10% 10% 10% Marketing 5% 15% 25% Research Grants 0% 15% 25% Research Recruiting 25 base projects 15 additional projects 25 additional projects Page 10

11 Expectations Revenue and Growth We have created a set of goals for the first four years after the release of the proprietary network. Projections are based upon a goal of 70% network resources devoted to commercial applications at the end of the 3rd year. With static research grants totaling 3,000,000 SPARC weekly, our goal is to sell 7,000,000 SPARC each week. We have factored in a conservative increase in price of the coin for each year, representing 33%, 25% and 20% growth in price respectively. The pricing is based upon Amazon AWS EC2 Elastic Compute Units, with the average miner being equivalent to a M4.2xlarge $0.40 USD per hour. The foundation takes 15% of revenue to put towards support and growth. Date End of Q End of Q End of Q SPARC Sold Weekly 1,000,000 3,000,000 7,000,000 SPARC Price $0.04 USD $0.05 USD $0.06 USD $40,000 USD $150,000 USD $420,000 USD $238 USD $893 USD $2500 USD 25% 50% 70% $0.09 USD/Hour $0.17 USD/Hour $0.24/Hour $65.70 USD $ USD $ USD Weekly Revenue Hourly Revenue Business Resources Miners Required2 Miners Earn Miners Monthly Pay 1 1 Business Resources represent resources used by revenue creating applications. 2 Miners Required represents the entire network including resources devoted to research. Page 11

12 Competitors We have two distinct competitor groups. Traditional supercomputer centers who run their own hardware, and other distributed computing networks. Supercomputer Centers Supercomputing centers are the current supply of power in the market. These centers are very expensive to run requiring their own hardware and support. As such we believe we can offer a 66% savings over traditional computing centers. This is primarily due to the cost of the hardware. Since our network relies on excess power using miners already existing machines we can run a more efficient operation. Other Distributed Networks We have identified Golem.network, SONM.io, and iex.ec. These three direct competitors are currently in the process of development and have not yet released versions of their products. We believe we will beat them to market, with our light-weight cross-platform application. Golem is a Polish startup, SONM from Russia, and Iex.ce has offices in France and China. We plan to be taking over the North American market for computing power. Although we have a home field advantage in North America, we are still internationally focused, and plan to release across various languages. Howey Disclaimer Due to the lack of legislation in the cryptocurrency economy, most online exchanges do not accept coins which are considered securities. As such we must fail the Howey test in order to be listed. Although we intend to use ICO funding to grow and scale the SPARC economy, by purchasing SPARC as a part of the crowdsale you agree that you are in no way expecting a return of profit. SPARC do not represent a share in the foundation; they are products of the foundation. We are in fact a Canadian company, however the same securities law has been endorsed as good law in Canada. For more reference please see: USA: SEC v. W. J. Howey Co (1946), 328 U.S Canada: Pacific Coast Coin Exchange of Canada v. O.S.C. (1977), 80 D.L.R. (3d) 529. Page 12

13 Team We are the founders of QuetzalPOS.com (2012), a cloud based point-of-sale system. Before that we were at Vivonet.com (POS) after DAS Software (1998, Solutions, E-commerce) was acquired (2008). We've been handling online transactions since 2002 and are naturally moving into blockchain technology. DOUGLAS STEWART - FOUNDER, CEO A serial entrepreneur this will be Doug's third company. His first successfully acquired, the second Quetzal is still growing. With 20 years of management he brings valuable experience and connections to the team. GREG AGNEW - FOUNDER, LEAD DEVELOPER Greg was a also founding member of the Quetzal four. Always at the forefront of technology, he has mastered Ethereum and full stack development. Leveraging the full force of Amazon's Lamda infrastructure he has ensured effective scaling. BRAD MARSHALL - FOUNDER, UI/UX DESIGN Brad has been working for Kingston's premier development studio, 14 theories, as a UX designer. He was a founder of Quetzal and is returning to the team for another round of excitement. Working with branding and UI design he will be polishing our products. EDDIE ROOSENMAALLEN - SYSTEMS ARCHITECT Eddie is a full-stack engineer with over a decade of experience in rich web apps on both physical and cloud deployments. He is a veteran teammate of ours, having worked together at both DAS and Vivonet. CAPT DANIEL DESJARDINS, PHD - RESEARCH ADVISOR A professor for the Department of Physics and Space Science at the Royal Military College of Canada, Capt Desjardins will be providing valuable knowledge about the needs of our researchers. MACK VAN ROSSEM - RESEARCH ADVISOR Former South Pole data center operator for the IceCube Neutrino Observatory Mack has worked slicing data and sending it over networks for use back home. He has a passion for reducing wasted computational power and improving our efficiency with this critical resource. Page 13

14 Glossary API Application Program Interface The API exposes specific subroutines to a third party while protecting the application from malicious interference. Blob Binary Large Object A Blob is a collection of binary data stored as a single entity. Block A blockchain block is a page in the ledger representing a finite amount of transactions to be added on top of the heap. Blockchain A blockchain is a distributed database that is used to maintain a continuously growing list of records similar to a ledger. They are typically manage by peer-to-peer networks collectively adhering to a protocol. BOINC Berkeley Open Infrastructure for Network Computing BOINC is an open-source network supporting volunteer distributed computing for research projects. Difficulty Level Difficulty is a measure of how difficult it is to find a hash below a given target. The target set is a numerical number represented in hexadecimal. The competing miners must find an acceptable hash of the required transactions below that number. Elastic Compute Unit Elastic Compute Units are abstraction of computer resources used by Amazon EC2. They represent a single thread with GHz of processing power. Hashrate Hashrate is the rate at which a computer can compute a hexadecimal representation of the required transactions and salt according to the desired algorithm. Page 14

15 PFLOP Peta-Flop A Peta-Flop is 1015 floating point operations per second. Used to represent computing power they are a primary mathematical operation for scientific research. Salt (cryptography) A salt is a random piece of data used to change the final hash when combined with otherwise static data in a hashing algorithm. SPARC Science Power and Research Coin are the tokens used to reward miners and purchase power on the KDS network. External Links Amazon AWS - Bitcoin - BOINC - Ethereum - Golem - Google Cloud Compute - Gridcoin - Iex.ec - OpenGL - Seti@home - SOMN.io - Top 500 Supercomputer List - Page 15