ICO Analysis: Hypernet
Hypernet technology allows personal computers, smartphones and even fridges with processors to connect with each other and become one completely ”parallel” supercomputer. The combined synergy between numerous devices can create calculating capacities which can rival every existing supercomputer.
With Hypernet it will be possible to lease your unused computing power to other users who want to use it and are willing to pay for it. Sellers register their devices in Hypernet and set the price for the time. Then, like traditional cloud services, buyers write a program and ask for hours of computing on a certain number of devices to run their program. Hypernet ensures that the buyer’s program will be divided and safely sent to each of the devices, processing arbitration between users.
Let us touch the topic of supercomputers in general.
Any supercomputer is just a just collection of computers. They are optimally connected and allow synergy between devices. However, these supercomputers are very bulky and require maintenance. The second development is a distributed supercomputer, which is based on the same concept, but devices are connected through the internet. Currently, supercomputers can only work on specific tasks. Hypernet intends to solve this problem by creating a network in which all operations will be performed much faster and in parallel. The resulting in Hypercomputer is not limited in its computing abilities in the way like other distributed supercomputers.
From a tech point of view, Hypernet is innovative because of several reasons:
- Computation works for almost all types of problems, while competitors in distributed computing have tried to approach the most straightforward class of programs called grid computations.
- An algorithm based on a distributed average consensus that was developed from scratch especially for dynamic, distributed and decentralized device networks.
This gives Hypernet the ability to run all classes of problems in parallel which results in an infinitely scalable network that is not limited to the cost and time required to create new data centers. This allows data reduction for massively-parallel sets of tasks.
Hypernet’s direct competitors are SONM, Golem and Constellation Labs.
Hypernet has three main elements, each with a separate roadmap:
Blockchain Resource Scheduler – The Hypernet Foundation will work with developers to create a schedule in the form of a graphical dashboard. Sellers will be able to use the scheduler to specify the conditions for using their equipment (when the equipment can be used, how much memory and disk should be available, and how much of their processor time they are willing to offer).
Hypernet Consensus Infrastructure – The Hypernet Foundation will work with third parties to develop an API library for distributed average consensus in a hyperlink with a lever.
Hypernet Executable Environment – To ensure security and interoperability in the system, it is expected that the Hypernet Foundation will create a sandbox environment for executing executable files to isolate them from the vendor’s hardware.
The project has a sophisticated token use structure.
STAKING: Buyers and sellers must stake HyperTokens to complete compute jobs.
REPUTATION: A user’s reputation increases by being a reliable and responsible compute provider and compute purchaser, and this reputation is permanently logged on the blockchain. A user’s reputation increases the likelihood of participating in compute jobs.
CURRENCY: HyperTokens are the transactional currency which enables the buying and selling of compute on the network.
AVAILABILITY MINING: Individuals can mine HyperTokens while waiting for compute jobs, by just being available in the lobby. This incentivizes users to join the network and make their devices available.
DECENTRALIZED GOVERNANCE / VOTING: Nodes participate in challenge and response and are incentivized for helping maintain the quality of the network, and weeding out bad actors.
There is no information on token distribution and use of funds available.
Below is a rundown of key team members.
- CEO and Founder
- Graduate of Stanford.
- Development of start-up products in the companies Ad Astra Rocket Company as plasma physicist
- LanzaTech as a chemical engineer.
- Candidate research at Stanford focused on advanced space motion from magnetoplasma missiles to expanded theories of gravity.
- Co-Founder and CTO
- Graduate of Stanford.
- Awarded a US Defense and Engineering Scholarship in the Department of Aeronautics and Astronautics, Stanford, for his thesis.
- Current research interests relate to fault-tolerant algorithms for distributed and exact computations and the application of optimal control methods for the training of stabilized neural network architectures.
- Co-Founder and CFO
- Graduate of Stanford
- Founded in 2013 a company of solar power electronics Dragonfly Systems with a successful exit when the company acquired by SunPower Corporation in 2014.
- Remains advisor to SunPower
- In 2009-2013 years. was a co-founder of the non-profit organization weJAYA, which dealt with the fight against poverty in West Timor.
The key advisors are listed below.
- Randall is a co-founder of Akamai Technologies, the global leader in Content Delivery Network (CDN) services. Randall is also the founder and CEO of JUMP Investors, a venture capital firm that also functions as his family office (notable investments include Google and Seagate.)
- Marc Weinstein is the Head of Research & Analysis at DNA Fund.
Below is the breakdown of risks and growth factors of the project.
- The competition in the sector is very large. Both from centralized systems like Amazon, and from decentralized (Golem) and those that are doing ICO at the moment (Ankr Network, Arpa, Teex) -1
- The team does not have much experience in parallel computing and blockchain experience -1
- No information on token metrics at this point -1
- WP technical assumptions are very questionable as the use of blockсkchain is not entirely understandable in view of the fact that the same system based on HyperLeger can do everything as effective -1
- The mainnet launch was scheduled to be released in August, but there are not new releases of development announcements at this point. -1
- hard cap of 15million is pretty reasonable for this space +2
- The team is a good one for a start-up. Young graduates of good universities and an interesting background +2
- Have some institutional support (The DNA fund) +2
- MVP is available +2
- There is a market for growth in the space +2
In general, the project can be assessed as average. The team is good but without much experience in the industry. The product is interesting, but there are delays in implementation, which for me is a red flag. 5 out of 10.
- Type: Utility
- Symbol: Hypertokens
- Platform: Ethereum/Own platform
- Crowdsale: June
- Minimum Investment:
- Price: TBA
- Hard Cap: 15 million USD
- Payments Accepted: ETH
- Restrictions Barred from Participating: TBA