Thursday, April 15, 2021
ST Aims to Embed its ToF into Every Power Switch
ST Micro publishes a use case of "Touchless Switch Using ST Time-of-Flight":
ON Semi and Ambarella Promote FaceID Access Solution
ON Semi publishes a promotional video "Smart Access Control with AR0237IR":
Gigajot Achieves 0.19e- Readout Noise in 16.7MP Sensor
IEEE EDL publishes Gigajot paper "A 0.19e- rms Read Noise 16.7Mpixel Stacked Quanta Image Sensor with 1.1μm-Pitch Backside Illuminated Pixels" by Jiaju Ma, Dexue Zhang, Omar A. Elgendy, and Saleh Masoodian.
"This paper reports a 16.7 Mpixel, 3D-stacked backside illuminated Quanta Image Sensor (QIS) with 1.1 μm-pitch pixels which achieves 0.19 e- rms array read noise and 0.12 e- rms best single-pixel read noise under room temperature operation. The accurate photon-counting capability enables superior imaging performance under ultra-low-light conditions. The sensor supports programmable analog-to-digital convertor (ADC) resolution from 1-14 bits and video frame rates up to 40 fps with 4096 x 4096 resolution and 600 mW power consumption."
Wednesday, April 14, 2021
Tesla Self Driving Becomes Camera-Only, Abandons Radar
CleanTechnica and Teslarati quote Elon Musk tweets that the oncoming Full Self Driving (FSD) Beta V9.0 will rely only on cameras and use no radars anymore:
Almost ready with FSD Beta V9.0. Step change improvement is massive, especially for weird corner cases & bad weather. Pure vision, no radar.
— Elon Musk (@elonmusk) April 9, 2021
Teslarati adds more Elon Musk's tweets:Remove
— Elon Musk (@elonmusk) April 9, 2021
When radar and vision disagree, which one do you believe? Vision has much more precision, so better to double down on vision than do sensor fusion.
— Elon Musk (@elonmusk) April 10, 2021
These things are best thought of as probabilities. There are 5 forward-facing cameras. It is highly likely that at least one of them will see multiple cars ahead.
— Elon Musk (@elonmusk) April 10, 2021
Smartsens Presents "Blue Ocean of Market Opportunities" for 3K Video
Smartsens unveils a "3K Line of Image Sensors" consisting of a previously announced 60fps SC500AI and a slower 30fps version of it named SC501AI.
3K Video Blue Ocean New Opportunity:
At present, 2K is the mainstream resolution of the image market. But in intelligent applications such as face recognition and intelligent detection, 2K resolution has gradually become unable to meet the market demand for high-definition video. However, in the process of many high-end video applications transitioning to 4K resolution, 4K video is limited by the transmission bandwidth. After the H.265 encoding and compression, the final rendering effect did not meet the application expectations for 4K images. Therefore, a blue ocean market opportunity for 3K video was born between 2K and 4K.
Ouyang Jian, Deputy General Manager of Smartsens, says, "With its continuous in-depth exploration of customer differentiation needs, Smartsens is keenly aware of the huge CIS Blue Sea market between 2K and 4K imaging and has launched a 3K-class product line to better meet the upgrade of end-customer application needs. In the future, Smartsens will continue to launch a number of different series of products to meet the growing application needs of various segments of the future HD video era."
Omnivision Announces Nvidia Drive-Qualified Automotive Sensors
BusinessWire: Omnivision uses Nvidia GPU Technology Conference to announce 2 new automotive sensors compatible with Nvidia Drive platform.
A new un-named 5MP image sensor is an automotive global shutter sensors for DMS applications. It also adds Nyxel NIR technology to achieve the world’s highest QE for DMS of 36% at 940nm wavelength.
As the industry’s first automotive image sensor with 140dB HDR, along with LED flicker mitigation (LFM) and high 8.3MP resolution, the new OX08B40 aims to front-view cameras, powered by an on-chip HALE (HDR and LFM Engine) combination algorithm.
Tuesday, April 13, 2021
Reuters about ON Semi Automotive Supply Chain
Reuters gives an example of automotive camera international supply chain:
"...consider a chip supplied by a U.S. firm for Hyundai Motor Co’s new electric vehicle, the IONIQ 5.
Production of the chip, a camera image sensor designed by On Semiconductor, begins at a factory in Italy, where raw silicon wafers are imprinted with complex circuitry.
The wafers are then sent first to Taiwan for packaging and testing, then to Singapore for storage, then on to China for assembly into a camera unit, and finally to a Hyundai component supplier in Korea before reaching Hyundai’s auto factories.
A shortage of that image sensor has led to the idling of Hyundai Motor’s plant in South Korea, making it the latest automaker to suffer from global supply woes that crippled production at most automakers including General Motors Co and Ford Motors Co and Volkswagen.
And the winding journey of the image sensor shows just how complicated it will be for the chip industry to both ramp up capacity to address the current shortage and re-invigorate U.S. chip manufacturing.
“Trying to reconstruct an entire supply chain from upstream to downstream in a single given location just isn’t a possibility,” David Somo, senior vice president at ON Semiconductor, told Reuters. “It would be prohibitively expensive.”
High Speed Counting SPADs
Phys.org: NIST publishes an Applied Physics Letters paper "Low-noise photon counting above 100 × 106 counts per second with a high-efficiency reach-through single-photon avalanche diode system" by Michael A. Wayne, Joshua C. Bienfang, and Alan L. Migdall.
"We demonstrate a method that allows a high-efficiency single-photon-avalanche diode (SPAD) with a thick absorption region (>10 μm) to count single photons at rates significantly higher than previously demonstrated. We apply large (>30 V) AC bias gates to the SPAD at 1 GHz and detect minute avalanches with a discrimination threshold of 5(1) mV by means of radio frequency interferometry. We measure a reduction by a factor of ≈500 in the average charge per avalanche when compared to operation in its traditional active-quenching module and a relative increase in >19% in detection efficiency at 850 nm. The reduction in charge strongly suppresses self-heating effects in the diode that can degrade performance at high avalanche rates. We show that the single-photon detection system maintains high efficiency at count rates exceeding 1e8 s−1."
Samsung to Manufacture Image Sensors at UMC
TheElec: Samsung has recently signed a contract for UMC to manufacture Samsung’s image sensors in Taiwan, according Taiwan Economic Daily.
As a part of the deal, Samsung is selling about 400 units of fab equipment to UMC to be installed at USMC's P6 fab at Nanke and used in the sensor manufacturing. UMC expects to produce 27,000 28nm wafers per month at the fab starting sometime in 2023.
The Taiwan Economic Daily says the P6 fab will mostly manufacture chips for Samsung.
Samsung sees high demand for its image sensors, but it has suffering from a shortage in production capacity. The new Samsung-UMC deal somewhat mirrors Sony-TSMC one, although Sony probably had different reasons for going to TSMC for advanced processes.
Monday, April 12, 2021
Subscribe to:
Posts (Atom)