Filament Chip resemble to me leonardo da Vinci Elicopter made of wood, but not so made of wood, the filament chip seem a very clean and precise project with respect of the standard with lot of intuition on the future really hard to imagine in a contest of the technology of the 1996.
Near 20 years are passed from when this wonderful project was introduced.
Let dicover it with a lens on the past,
when you read it imagine that fringe internet in 1996 consisted of pentium 90Mhz and analogic modem to connect the internet, email address was so strange an unfamiliar like bitcon addresses now...
...In the spring of 1996, Professor Neil Gershenfeld posed the challenge: design an 'IP Lite' chip to make it easy to connect something as simple as a light switch to a computer network.
"Early in the process, we decided on a few basics on which to base subsequent design decisions."
- The fundamental goal is to specify a chip (or a small set of chips) that makes it cheap and easy to connect to existing networks. Our test for a characteristic application was to ask ourselves "what would be required to sense the state of a lightswitch?"
- The core Filament Chip will be independent of the Link Layer. A specific link layer (10BaseX, fiber, IrDA etc) may be implemented as a separate chip or integrated on the same chip.
- The "host" is assumed to be a small, relatively slow chip with limited RAM, such as a PIC processor. The Filament Chip must take on any real-time requirements for buffering datagrams and sending acknowledgements.
The Filament Chip is intended to provide simple network connectivity in return for a small investment in hardware and design. With this in mind, the group pondered several network applications of the future, including:
- Put a toaster on the Ethernet.
- Put a VCR on the net, to be programmed via TV Guide(TM) web interface.
- Implement a home security systems using kerberized protocols.
- Create sensor gateways in furniture.
- Create a Universal remote control that's actually somewhat universal.
- Create "smart" appliances that can call the factory for diagnostics, firmware upgrades, and service requests: your dishwasher calls the Maytag man.
- Smart vending machines which report inventory on demand.
- Interphones: telephones connected by computer networks
- NTP (Network Time Protocol) *everywhere*: no more blinking clocks, create a gateway to your Swatch.
- Smart HVAC ==> energy conservation with various optimizations.
Here a table of an hypotetical 14 pin configuration:
|2||OSCI, OSCO -- crystal oscillator input and output pins.|
|4||mTxD, mRxD, mCLK, mCTRL -- four wire sync/async interface to a link layer chip.|
|4||hTxD, hRxD, hCLK, hCTRL -- four wire sync/async interface to the host.|
|1||INT -- interrupt line to host (could share a pin with hCTRL)|
|1||RESET -- when held low at power up, forces a reset of the chip to its default values.|
As a bare minimum, the Filament Chip must support IP, UDP, and ICMP (ping) protocols. For IP, the chip will neither generate nor accept fragmented packets.
In addition to these basic protocols, we assume the Filament chip supports a basic "discovery" protocol, such as BOOTP or DHCP (preferred). And layered atop UDP, we will assume SNMP for a simple communication model.
The chip will work in one of three primary modes, Discovery Mode, Datagram Mode or Management Mode.
The need for the Filament Chip was first articulated by Professor Neil Gershenfeld, director of the Physics and Media group, and co-director of the Things That Think research consortium. The entire Filament Chip effort owes its existence to Neil.
Full article here .