Zero instruction set computer
In computer science, zero instruction set computer (ZISC) refers to a computer architecture based solely on pattern matching and absence of (micro-)instructions in the classical sense. These chips are known for being thought of as comparable to the neural networks, being marketed for the number of "synapses" and "neurons".[1]
The acronym ZISC alludes to reduced instruction set computer (RISC).
History
ZISC is a hardware implementation of Kohonen networks (artificial neural networks) allowing massively parallel processing of very simple data (0 or 1). This hardware implementation was invented by Guy Paillet, developed in cooperation with the IBM chip factory of Essonnes, in France, and was commercialized by IBM.
Design
The ZISC architecture alleviates the memory bottleneck by blending pattern memory with pattern learning and recognition logic.
Their massively parallel computing solves the "winner takes all problem in action selection" by allotting each "neuron" its own memory and allowing simultaneous problem-solving the results of which are settled up disputing with each other.[2]
Applications and controversy
According to TechCrunch, software emulations of these types of chips are currently used for image recognition by many large tech companies, such as Facebook and Google. When applied to other miscellaneous pattern detection tasks, such as with text, results are said to be produced in microseconds even with chips released in 2007.[1]
Junko Yoshida, of the EE Times, compared the NeuroMem chip with "The Machine", a machine capable of being able to predict crimes from scanning people's faces, from Person of Interest (TV series) describing it as "the heart of big data" and "foreshadow[ing] a real-life escalation in the era of massive data collection".[3]
See also
- Content-addressable memory
- One instruction set computer (OISC)
- TrueNorth
References
- Lambinet, Philippe. "The Ongoing Quest For The 'Brain' Chip". TechCrunch.
- Higginbotham, Stacey. "Make way for more brain-based chips". Gigaom.
- Yoshida, Junko. "NeuroMem IC Matches Patterns, Sees All, Knows All". EE Times.
External links
- US Patent for ZISC hardware, issued to IBM/G.Paillet on April 15, 1997
- Image Processing Using RBF like Neural Networks: A ZISC-036 Based Fully Parallel Implementation Solving Real World and Real Complexity Industrial Problems by K. Madani, G. de Trémiolles, and P. Tannhof
- From CISC to RISC to ZISC by S. Liebman on lsmarketing.com
- Neural Networks on Silicon at aboutAI.net
- Zero instruction set computer at Curlie