[ad_1]
Barcelona, Spain – (BUSINESS WIRE) –(MWC, Hall 1, Stand 1C80 / in co-location with Egis Technology)– AIStorm,
an innovative high-performance AI-in-sensor processors, introduced
a family of semi-custom solutions for mobile phones, IoT and advanced
Driver Assistance Systems (ADAS).
David Schie, CEO of AIStorm, sees a huge opportunity to outperform approaches
need to scan the sensor data, a step that introduces latency and
the potential to miss important information.
"Many industry players are focusing on deep GPU / NPU-based technologies.
solutions to speed up the AI at the edge. We believe that such solutions
are not compatible with real-time processing, power and low cost
requirements of these applications, "said Schie." Today, we are
reshape the landscape by allowing complete solutions including a
Sensor, Analog Antenna and Low Cost Artificial Intensity Treatment, and
even for the smallest form factors – without the need to digitize inputs
The data. "
AIStorm initially targets two major markets, mobile / wearable and ADAS:
-
AIStorm IoT Vision / IoT Waveform Solutions: Made for
imaging and HID / biometric applications in mobile phones, cameras,
wearables and IoT applications, AIStorm offers AI-in-sensor solutions
for tasks such as fingerprint detection, gesture control,
Heart-based monitoring and identification, presence detection, facial treatment
recognition, voice input, headset, drone imagery, image stabilization,
and security cameras and intersection. Custom optics or high CMRR PGA
Solutions are proposed integrated with semi-personalized variants of
AIStorm High Performance Analog Artificial Intelligence Engines. -
AIStorm ADAS Solutions: Designed for advanced driver badistance
AIStorm presents the coupled CIS, SPAD / SiPM and SiGe sensors
to its AI badog engines, which eliminate the need for scanning
and allow continuous processing of incoming data. These systems
outperform ADC solutions, which can easily miss or distort
information and difficulty in providing selective data in real time
harvest. Solutions to NIR, 1550nm and CIS are proposed, as well as
mirror, polarization, high voltage drive circuit, image stabilization, and
image processing solutions. Other solutions include gestural control,
eye monitoring, voice and wavelet failure detection.
The direct use of sensor data, without scanning, allows the use of real-time data.
Processing
AI systems require that information be available in digital form before
can process data, but the sensor data is badog. Processing this digital
the information requires advanced and expensive GPUs / NPUs that are not suitable
for mobile devices as they require continuous scanning of the inputs
data, which consumes a lot of energy and inevitably introduces
scan time (latency). AIStorm removes these disadvantages by
processing sensor data directly in its native badog form – in real time
time – using load domain processing instead of GPU / digital processing.
"While the treatment of digital AI seems to be reaching its limits, the next generation
In-depth learning systems will require badog AI cores. With our recent
investment in AIStorm, we affirm our belief that the time has come to
built-in AI-in-sensor badog technology, and that AIStorm technology is
the right solution, "said Dr. Avi Strum, senior vice president and general manager of the sensor business.
TowerJazz unit. "The combination of the rich images of TowerJazz and
non-imaging sensor technologies with artificial intelligence from AIStorm
This core will bring unique solutions to the industrial, commercial and
automotive markets. "
About AIStorm
AIStorm is the pioneer and leader in AI sensor processing, which
eliminates the latency, power and costs badociated with digital
GPU / NPU-based implementations at the periphery. AIStorm is headquartered in
Silicon Valley and led by a team made up of veterans of the industry
responsible for the development of more than a thousand products, as well as
for the income statement and the significant revenue growth of major semiconductors
companies. For more information, visit https://aistorm.ai
or write to us at [email protected].
Source link
