Ether2
Startup/business
California,
United States
http://www.ether2.com
Ethernet's "listen and send" (CSMA/CD 802.3), best-effort signaling method introduced the foundational problem that created the need for the Cisco type switch/router solution. Furthermore, IP (Internet Protocol) was designed as a low cost alternative to private carrier class networks. It was never meant to carry everything, and router manufacturers would not exist if not for the inability of the underlying Ethernet to connect multiple network nodes to each other directly, or to transmit data efficiently over telephone company networks without an overlay of Ethernet network gear (modems and routers). Every home and business customer, therefore, pays for two networks in order to move one packet of data. For the lack of a better alternative, we are all forced to plug in asynchronous Ethernet products on top of the synchronous phone and cable company lines. After nearly 30 years of Ethernet becoming the dominant MAC, it still has delays from dropped ackets, back-off periods and retransmissions, keeping demand for hi-def content and voice quality unsatisfied.
Ether2, therefore, proposes a technology that is not just the next generation of Ethernet, but eliminates the need for the Ethernet overlay altogether. Our mission is to commercialize this technology which has the ultimate potential to eliminate the networking problems of the last 25 years that gave rise to centralized hardware that is superimposed on top of the local carrier, business and consumer networks in order to move data packets efficiently.
An Ether2 network’s central component can be a fiber, copper or wireless “dumb” hub whose chief responsibilities include monitoring the global queue, granting reservations, resolving contention for simultaneous requests, and broadcasting the data to one, to many or to all. Our switch’s size is unlimited as to the number of nodes (servers, PCs, mobile units, or wireless/mesh access points) it can serve. Low latency, high throughput, QoS, security, and superior hand-offs in ad hoc networks are the key features of our universal MAC (Medium Access Control), which can run disparate technologies (CDMA or TDMA, Packet or Circuit Switched, Synchronous or Asynchronous) at the same time.
We operate under an exclusive license with IIT (Chicago) and UPC (Barcelona). More than $250K has been invested in our chip and printed circuit board fabrication, and we seek $700K in additional funding in which to achieve proof of concept. Phase II funding of $1.2M will be required to complete the prototype, and Phase III funding of $3M in order to achieve commercial readiness with a wireless application.
Our initial application will be a wireless offering for the home, business or data center via USB and an Ether2 hub. After significant market penetration we will shift to an OEM model, offering embedded wireless, wireless optical, and wired (RJ-45 PHY) ports. Upon carrier adoption, the whole of the Internet could be converted to a distributed architecture, enabling wired and wireless carriers true convergence. Meaning, all forms of network communications would have direct access to the Layer 2 MAC without MPLS (Multiple Protocol Label Switching) sitting between Layer 2 and 3. This will eliminate congestion from streaming and would also offer a high-speed last mile solution over the carriers’ old telephone lines, new fiber runs, or wireless to the home initiatives. Additional benefits include multiple levels of QoS, embedded security, short messaging, and Layer 2 multicast/broadcast. Finally, Ether2’s Windows/Linux drivers will be cloaked as Ethernet drivers so that existing applications will migrate seamlessly. We will offer auto-sensing for legacy Ethernet nodes and bridging to Ethernet switches, as well as an easy migration by running Ethernet and Ether2 concurrently on the same Ether2 network.
Our wireless hub and a wireless USB device will be sold through direct sales and VARs. Once commercialized, Ether2 technology will spread quickly by offering free developer kits to a global open source developer community. In sharing all but our chip core, we will give rise to our own competitors who will either license our chip or try to copy it (with a very long R&D curve). Drivers will be wide open and customizable, and 50% of the chip can house other System on Chip (SoC) applications.
We are the near-IDEAL OPTIMIZATION of the channel, with mathematically predictable transmissions and none of Ethernet's drawbacks. There is no other UNIVERSAL network protocol that can theoretically reduce latency to the actual propagation delay without over-provisioning, or share nearly all of its bandwidth evenly among its users! As with any disruptive technology, the barrier to entry is the market’s complacency with the status quo and, in our case, the carriers’ long term Ethernet planning. Another barrier to entry is that we are not conforming to an IEEE standard, though (TCP/IP) will not know that, nor would any other protocol conform to a standard upon being invented and developed.
Final note: We know our website needs an update, but this is not a priority at this time.
Business model
Ether2's growth model can best be thought of as following the path of Linux, but at the network layer. We cite Linux in terms of being an open source technology which has severely eroded a monopolistic (Microsoft) server operating system market. However, since our solution is hardware-based, a first generation wired and wireless product line will be launched. These products will have the potential to give rise to new companies on every continent, as opposed to going head to head with Cisco and the like. In addition to these new companies who will license or attempt to copy our chip, revenue will ultimately be driven through OEM licencing to device manufacturers. Application consulting and Ether2 developer community events will also drive revenue.
Our current' prototype incorporates a daughter card which had to be designed from scratch, as there is no standard development board which does not expect to see an Ethernet chipset. The boards were then stuffed with a Xilinx FPGA core. Phase I funding will complete the testing and driver development needed to achieve proof of concept. Phase II will see the wired hub development, while we move from FPGA to ASIC, along with the conversion to an Ethernet PHY. Phase III will see the package move to commercial readiness with a wireless hub and wireless NIC via USB for a home or data center solution.
Competitive advantage
Ether2 essentially offers a game changing enabling technology, and this comes in the form of a universal network protocol for any type of carrier, broadcaster or wireless play. No other data packet transport method can claim to unify separate protocols for video content and data content, without engaging in the massive over-provisioning of Ethernet switches, which is the solution of the day. For nearly 30 years, this switched Ethernet legacy solution has always led to the networking bottlenecks of tomorrow, and the carriers keep paying for more network bandwidth upgrades, but what they do not realize is that they just keep making the same dirt road wider.
Ether2, therefore, can allow traditional carriers to re-invent themselves at a time when nobody really needs the phone company for dial tone anymore. In a wired or wireless environment, we would allow anyone to broadcast hi quality video to one, to many or to all the users on the network. Consumers would benefit because new media content providers could offer better quality than traditional broadcast networks. Of course, at this level, the wireless carrier would be offering a "last mile" solution that could compete with fiber to the home from traditional carriers.
Today's fiber switches are limited to 128 ports. The Ether2 solution has no upper bound. Aside from that, no other company is even thinking about technologies that could lead to the realization of the "one machine" concept, in which a series of cascaded hubs would create an environment in which there is only one virtual hop between any two nodes on the planet. So while physical networks would remain largely the same, the virtual network would be completely flat, like a single string that goes in any direction to connect every server and pc. This would eliminate the need for geographically dispersed server farms because latency would be reduced to propagation delay, over any distance to any number of users.
Ether2 will also offer backwards compatibility for legacy Ethernet nodes, so that they can transmit at the same time as the rest of the network. Of course, there will be a performance hit for reserving Ethernet time slots on a busy network, but for networks with less traffic, Ethernet can be interleaved between the standard data packets for a smooth migration path.
Finally, because of Ethernet's limitations, the latest competitive research has been in an attempt to make Ethernet synchronous. It seems that Ethernet is trying to use synchronous queuing/polling to provide greater efficiency in 802.11 and 802.16, however, the results have not been stellar because of Ethernet's asynchronous nature. See http://scenic.princeton.edu/research/researchCSMA.html
Management Bio
Cofounder & CEO David Dietrich knows about moving massive amounts of data. Mr. Dietrich has led major network and software projects with staff exceeding 100 employee engineering teams; in fact, if you have ever booked an airline reservation on the Internet, chances are that you have used his systems, which are responsible for more than 10 million transactions annually. Mr. Dietrich has also been a senior network architect for facial recognition security systems in China’s airports where he learned to speak Chinese. His professional credits also include positions as technical lead for the development of database management systems for companies such as McDonnell Douglas and Houghton Mifflin, and he also directed engineering teams through a full software development cycle for American Express, Travelocity and Navigant International. Additionally, Mr. Dietrich is a past president of the American Marketing Association, and has served as a professor of E-commerce at San Bernadino State University, which is also where he received his MBA in Marketing. Today, as CEO of Ether2, Mr. Dietrich utilizes all of his skills, both in marketing and raising funds, as well as in assisting the engineering team by writing Ether2’s Linux drivers, which is only one of several programming codes in which he is extremely proficient. Based in Southern California, he looks forward to building the premier Layer 2 networking company and steering the ship to the realization of Dr. Graham Campbell’s vision…routing without routers and switching without switches.
Dr. Graham Campbell, Cofounder, Ether2 Corporation -Currently
serving as Technical Advisor & Professor of Computer Science &
Engineering (Ret.), Illinois Institute of Technology, Dr. Graham Campbell
received the BScEE and MScEE degrees from the University of Manitoba in Canada.
After working in industry in Canada he emigrated to the USA and established
with a classmate a company providing consulting services related to control
systems to NASA, the US Air Force, and the U.S. Army. Subsequent to this he
joined Penn State University where he taught and also directed Computer Science
activities at the 20 Commonwealth Campuses of the Penn State system while
completing his PhD in Computer Science. He returned to Canada in the mid 70s
for a four-year stint as Science Advisor to the Premier of Manitoba. He was a
Professor of Computer Science at the Illinois Institute of Technology for over
20 years, where he was responsible for the courses related to data
communications. In working with
fourteen students who completed their PhDs under his direction, Dr. Campbell
and his team invented the Distributed Queue Switch Architecture (DQSA).
Scot Stockton,
Engineer, Ether2 Corporation -Mr.
Stockton spent 17 years as a chip designer at Northrop Grumman. His specialties include ASIC
(Honeywell), FPGA, Spacecraft attitude control, On-board Data Storage, Uplinks,
Downlinks, & Crosslinks, High-bandwidth Data Routing, Discrete Analog, RF,
& Digital, Board Design, Payload Slice Design, Sensor Tracking, Command,
Control and Telemetry processing, Communications, Multi-processor applications,
Modulators and Demodulators, A/D & DAC circuitry, VHDL, UNIX, Windows,
Synopsys Design Compiler, SPICE, C, perl, MS Office suite, Planning,
Scheduling, Budget Oversight and Document Writing.
Roger Thorpe, Engineer, and CEO of Advanced Architectures
(A2), an Orange County firm designing high-performance computer systems, subsystems and components. A2’s
strengths include the conceptual design and architecture of complete systems
including custom DSPs and CPUs. A2’s
design experience includes implementing structured design methodologies,
performing HDL modeling simulation, performance analysis, logical design,
packaging and detailed implementation.
Dr. Luis Alonso, PhD. Director of Signal Theory and Communications, Polytechnical University of Catalonia, Spain. Dr. Alonso received an Engineer of Telecommunications degree from the Universitat Politécnica de Catalunya (UPC), Spain, in 1997. In 1998 he joined the Escola Tècnica Superior d'Enginyers de Telecomunicació de Barcelona (ETSETB), Spain, where he started teaching pre-graduate students while beginning his PhD preparation. At that time, he discovered the ideas of Dr. Graham Campbell about an amazing near-optimum access protocol suitable for any kind of communications systems. He decided to contact with Dr. Campbell and focus his research work based on this surprisingly good idea. Indeed, his PhD consisted in the proposal, modeling and analysis of the adaptation of Dr. Campbell’s protocol to a mobile wireless communications system based on CDMA technology. This PhD thesis was awarded the cum laude qualification in 2001, when Luis Alonso reached the PhD degree. He has participated in several research programs, networks of excellence, COST actions and integrated projects funded by the European Union and the Spanish Government, always working on the design and analysis of different mechanisms and techniques to improve wireless communications systems. He has been publishing several works (in international congresses, journals and magazines) based on the Distributed Queuing principle, trying to show all the benefits that this key idea may provide for all kind of wireless communications systems, when it is embedded into the general Radio Resource Management entities. He has been showing also that QoS features and Cross-Layer optimization are especially easy to fit into the DQSA-like architecture. Furthermore, he is currently supervising three doctoral students who are working in different aspects of applying the Distributed Queuing.
Gary Bahadur is a Cofounder of Ether2 Corporation, in which his responsibilities include boot strapping, marketing, and product development. Prior to Ether2 he was a Co-Founder and served as Chief Information Officer of $20 million vulnerability risk management company Foundstone with 5 offices around the United States and 1 office in Singapore. Foundstone was sold to McAfee Inc in September 2004 for an all cash deal of $86 million. Foundstone offered a unique combination of proprietary software, services, and education to help customers mitigate and manage the digital security risks inherent in doing business. As co-founder of Foundstone Inc, Gary’s duties ranged beyond traditional CIO requirements. Key activities included growing Foundstone Inc from 6 to 130 employees with areas of responsibility such as business development, marketing, consulting and training. Prior to starting Foundstone Gary worked at Ernst & Young assisting in developing information security practice. He performed numerous penetration studies and network reviews covering various firewalls, Windows NT and Novell networks, Web servers and Internet connectivity, SAP security in UNIX and Windows NT environments, routers, and different flavors of UNIX. Gary was responsible for client service, developing methodology, developing content for Vulnerability Service of the eSecurity Solutions group, assisted in developing and teaching the “Extreme Hacking – Defending Your Site”. Prior to Ernst & Young, Gary was a Senior Consultant at Price Waterhouse LLP, responsible for security team, conducting security reviews of Global 2000 clients and developing methodologies for security reviews. Gary holds a Bachelor of Science degree in Finance and Information Systems from New York University.
IP and defensibility
In 2003,
Buddenberg, of the US Navy, writes a paper entitled “Radio WAN Media Access Protocol” in which
he outlined five requirement for optimum next generation networks. The DQSA technology, including a family of patents licensed by Ether2 from Illinois Institute of Technology, satisfies all of them.
Requirements for a radio-WAN MAC.
These requirements would pertain to any solution, not just a satellite communications one.
1. The primary requirement is to discipline network nodes so they transmit one at a time. The solution space becomes restricted to scheduling and polling algorithms by recognition of this requirement.
2. Provide a means for new nodes to enter the network segment. Most prospective solutions envision a second queue for management of this problem.
3. Support multicast. This definition of multicast is slightly different than the layer 3 definition (delivery to multiple platforms for price of a single transit through each router).
4. Support urgent access for priority traffic. This is one facet of the Quality of Service control problem.
5. Provide synchronous access for those customers with requirements for determinism.
Follow Ether2 on Vator.tv
