The smallest computer in the world, an energy efficient chip and other innovations for IoT
Today we have prepared a digest with the new "chip design" for IoT. We’ll tell you about new devices for data encryption, the smallest IBM computer in the world and NVIDIA solution, which simplifies the integration of deep learning systems into microprocessors.
/ photo Santi CC
At the IBM Think 2018 conference in March, the company introduced the smallest computer in the world. Its dimensions are 1x1 mm, which is even smaller than a grain of salt.
')
The computer has a processor with several hundred thousand transistors, a COS, a solar power supply system, and a communication module with LEDs and a photo detector. By power, the microcomputer will be equal to the processor with the architecture of the x86 of the 90s.
IBM was told that the new microchip will be used in blockchain technologies: it will serve as a source of data for blockchain applications. For example, for logistics companies, this solution will help detect fraudsters in the supply chain.
This makes it possible to track the origin of the goods. The low cost of chip production (about 10 cents) will allow massively embedding chips, for example, in electronic equipment, so that buyers can track where the goods came from and verify its quality. IBM calls its chip “crypto-anchor” (crypto-anchor), protecting data from theft and alteration.
In addition, the microchip will be able to perform simple tasks for AI systems, for example, to classify the data provided.
Release dates have not yet been named, but it is known that developers are already testing the first prototype. ZDNet claim that the microchip will appear on the market in a year and a half. In TechCruch predict the emergence of new within 5 years.
Scientists from the South Korean company SK Telecom have developed a microcomputer capable of generating truly random numbers. Such generators have already been created earlier and are even used in the operation of cryptographic systems. However, the Korean company was the first to translate this idea into a 3x5x1 mm chip ( LxWxH).
A tiny random number generator will be used in IoT devices to guarantee the protection of encrypted data during its transfer to other devices.
The device uses the phenomenon of quantum shot noise (quantum shot noise). The chip's LEDs emit photons that “bounce off” the inner walls of the device. The built-in CMOS matrix picks them up, and the pulses generated by it are already transmitted to the randomness-extraction algorithm.
SK Telecom and Nokia first demonstrated this technology in action last year. During the experiment, SK Telecom server generated encryption keys and transferred them to the Nokia 1830 fiber optic switch.
The exact cost of the device is not known, however, Sean Kwak, head of the laboratory of quantum technology at SK Telecom, says that it will amount to "a few dollars."
The Massachusetts Institute of Technology (MIT) has developed an energy efficient microchip that consumes 400 times less energy than software implementations of public key encryption. In this case, the device works 500 times faster.
Like most public-key cryptosystems, the chip uses elliptic cryptography techniques. At the same time, it is able to work with any elliptic curves, and its blocks, “responsible” for modular arithmetic, can handle numbers up to 256 bits long (classical systems work with 16 or 32 bit values). The Datagram Transport Layer Security ( DTLS ) protocol, which is responsible for processing encrypted data, is “wired” into the chip, which reduces the amount of memory required for its operation.
Testing and specific plans for using the device at MIT have not yet been reported.
/ photo Fritzchens Fritz PD
As part of the collaboration announced by NVIDIA President Jensen Huang, the two companies decided to integrate the open architecture of the NVIDIA Deep Learning Accelerator (NVDLA) into the Arm Project Trillium platform for machine learning. The joint project is designed to facilitate and accelerate the introduction of deep learning systems in mobile and IoT devices.
NVDLA is an accelerator for deep learning systems, having an open architecture and being built on the basis of the NVIDIA Xavier processor. NVDLA is based on NVIDIA's powerful developer tools (these are drivers, libraries, SDKs), among which new versions of the TensorRT programmable deep learning accelerator will soon appear.
As for the Arm processor, it is specifically “ sharpened ” for working with machine learning systems. It performs more than 4.5 trillion operations per second (on mobile platforms), and this number may increase by 2–4 times during its “overclocking”.
Companies hope that together these solutions will help chip manufacturers and developers to simplify the integration of AI systems into processors for IoT devices and provide the market with affordable products that support machine learning.
PS Materials from the First Corporate IaaS Blog:
PPS Selection of posts from our blog on Habré:
/ photo Santi CC
The smallest computer in the world from IBM
At the IBM Think 2018 conference in March, the company introduced the smallest computer in the world. Its dimensions are 1x1 mm, which is even smaller than a grain of salt.
')
The computer has a processor with several hundred thousand transistors, a COS, a solar power supply system, and a communication module with LEDs and a photo detector. By power, the microcomputer will be equal to the processor with the architecture of the x86 of the 90s.
IBM was told that the new microchip will be used in blockchain technologies: it will serve as a source of data for blockchain applications. For example, for logistics companies, this solution will help detect fraudsters in the supply chain.
This makes it possible to track the origin of the goods. The low cost of chip production (about 10 cents) will allow massively embedding chips, for example, in electronic equipment, so that buyers can track where the goods came from and verify its quality. IBM calls its chip “crypto-anchor” (crypto-anchor), protecting data from theft and alteration.
In addition, the microchip will be able to perform simple tasks for AI systems, for example, to classify the data provided.
Release dates have not yet been named, but it is known that developers are already testing the first prototype. ZDNet claim that the microchip will appear on the market in a year and a half. In TechCruch predict the emergence of new within 5 years.
True random number generator from SK Telecom
Scientists from the South Korean company SK Telecom have developed a microcomputer capable of generating truly random numbers. Such generators have already been created earlier and are even used in the operation of cryptographic systems. However, the Korean company was the first to translate this idea into a 3x5x1 mm chip ( LxWxH).
A tiny random number generator will be used in IoT devices to guarantee the protection of encrypted data during its transfer to other devices.
The device uses the phenomenon of quantum shot noise (quantum shot noise). The chip's LEDs emit photons that “bounce off” the inner walls of the device. The built-in CMOS matrix picks them up, and the pulses generated by it are already transmitted to the randomness-extraction algorithm.
SK Telecom and Nokia first demonstrated this technology in action last year. During the experiment, SK Telecom server generated encryption keys and transferred them to the Nokia 1830 fiber optic switch.
The exact cost of the device is not known, however, Sean Kwak, head of the laboratory of quantum technology at SK Telecom, says that it will amount to "a few dollars."
Energy efficient chip for IoT cryptosystems from MIT
The Massachusetts Institute of Technology (MIT) has developed an energy efficient microchip that consumes 400 times less energy than software implementations of public key encryption. In this case, the device works 500 times faster.
Like most public-key cryptosystems, the chip uses elliptic cryptography techniques. At the same time, it is able to work with any elliptic curves, and its blocks, “responsible” for modular arithmetic, can handle numbers up to 256 bits long (classical systems work with 16 or 32 bit values). The Datagram Transport Layer Security ( DTLS ) protocol, which is responsible for processing encrypted data, is “wired” into the chip, which reduces the amount of memory required for its operation.
Testing and specific plans for using the device at MIT have not yet been reported.
/ photo Fritzchens Fritz PD
Deep learning in IoT: a joint project of NVIDIA and Arm
As part of the collaboration announced by NVIDIA President Jensen Huang, the two companies decided to integrate the open architecture of the NVIDIA Deep Learning Accelerator (NVDLA) into the Arm Project Trillium platform for machine learning. The joint project is designed to facilitate and accelerate the introduction of deep learning systems in mobile and IoT devices.
NVDLA is an accelerator for deep learning systems, having an open architecture and being built on the basis of the NVIDIA Xavier processor. NVDLA is based on NVIDIA's powerful developer tools (these are drivers, libraries, SDKs), among which new versions of the TensorRT programmable deep learning accelerator will soon appear.
As for the Arm processor, it is specifically “ sharpened ” for working with machine learning systems. It performs more than 4.5 trillion operations per second (on mobile platforms), and this number may increase by 2–4 times during its “overclocking”.
Companies hope that together these solutions will help chip manufacturers and developers to simplify the integration of AI systems into processors for IoT devices and provide the market with affordable products that support machine learning.
PS Materials from the First Corporate IaaS Blog:
- How to seamlessly migrate to the cloud: 9 useful tips
- Cloud IT infrastructure: features of international projects
- Cisco UCS B480 M5 Blade Server: unboxing
PPS Selection of posts from our blog on Habré:
Source: https://habr.com/ru/post/353284/
All Articles