Hey Kachoo, preaching to the choir here my friend, BIG Wednesday is upon us, and all of us true believers are going to be basking in the glory of the green room, or living in the Maldives, whatever I don't mind.
www.eenewseurope.com
hase II
Absolute BrillianceI often go back to the Renesas tape out Ann to reread its context.
Key question for me is the following couple of paragraphs plus the common theme we are now seeing with 22nm both through Renesas in Dec and by BRN with GF...
They way I read the below is that Renesas have had someone come to them and request the tape out and device but open to interpretation I guess.
The other part is that Renesas will let the mkt decide on something like this as to the uptake and then decide if they will bring in house for control or just continue external 3rd parties to run with it.
Renesas tapes out spiking neural network chip
Renesas is taping out a chip using Brainchip's spiking neural network (SNN) technology and expanding internal and foundry capacity
“Now you have accelerators for driving AI with neural processing units rather than a dual core CPU. We are working with a third party taping out a device in December on 22nm CMOS,” said Chittipeddi.
The take up of the technology depends on the market adoption, he says.
“We want to see where the market reception is the highest, that is what determines whether we bring things in house or through a third party.”
Now...another recent article I was just reading is about GF and quite interesting group of names discussed including BRN.
Be nice to get in with Purdue as well. Intel in there and our own K Carlson published papers on STDP in 2011 via Purdue.
In a Series of Agreements, GlobalFoundries Buoys Chip Supply From Home
3 days ago by Biljana Ognenova
Tracking the goals of the CHIPS+ Act, GlobalFoundries is strengthening the U.S. semiconductor ecosystem with several partnerships and acquisitions.
The U.S. semiconductor manufacturing industry is beginning to reap the benefits of the CHIPS+ Act passed in July 2022. In a series of collaborations and acquisitions, GlobalFoundries (GF) is progressing toward one of the act's goals: to strengthen a U.S.-based semiconductor supply chain and reduce dependence on Asia-based manufacturers. Throughout the chip shortage, GF claims it has remained profitable by steadying productivity, containing cost, differentiating products, and doubling down on supply chain security.
A comparison of GF's and TSMC's EBIT margins. Image courtesy of Seeking Alpha
GlobalFoundries attributes its upward profitability to its strategic approach to chipmaking and its local partnership with key industry sectors, including automotive, memory, and computing as well as education.
GlobalFoundries Partners With GM, Renesas, and Purdue
In a major win for GF in the automotive sector, the foundry signed an agreement with GM on Feb. 9, 2023, to dedicate a manufacturing corridor at GF's upstate New York facility for GM’s key chip suppliers.
GF also secured more production capacity by acquiring Renesas' NVM resistive RAM technology. Specifically, GF now owns Renesas' proprietary Conductive Bridging Random Access Memory (CBRAM) technology, a low-power memory solution built for home and industrial IoT and mobile devices.
GF has secured wins in the education and R&D sector as well, partnering with Purdue University on a semiconductor education program. Purdue was among the first institutions to use the funds from the CHIPS+ Act to finalize a university-level semiconductor program. Now, with this new collaboration, Purdue's staff and students will use GlobalFoundries' facility and resources to create innovative, interdisciplinary solutions for advanced semiconductors and microelectronics. The partnership between GF and Purdue will provide next-generation professionals with hands-on expertise overseen at GF Lab.
GF Supports Next-gen Vision and Computing Technologies
GF is also championing the computing sector, namely with BrainChip's Akida neuromorphic chip built on 22 nm fully depleted silicon-on-insulator (FD-SOI) technology.
Akida architecture. Image courtesy of BrainChip
BrainChip is the world’s first company to develop ultra-low-power, event-based, neuromorphic AI IP to be used for always-on sensor applications. The AKD1500 chip was built on GF’s low-leakage FD SOI platform, promising an array of applications that don't overload the CPU.
Vision sensor specialist Oculi also recently announced a strategic partnership with GF, commissioning the foundry to manufacture its single-chip, intelligent software-defined vision sensor. The new sensor will be based on GF’s 55LPx, a platform that supports RF, ultra-low power, embedded NVM, and high-voltage BCDLite (a process technology).
Expanding U.S.-based Semiconductor Production
While GlobalFoundries relies on five-year agreements to ship chips from storage facilities in Dresden and Singapore, the foundry also plans to expand three U.S. locations, including one in Vermont and two in New York. Specifically, GF is broadening the scope of existing facilities to make 12 nm, 28 nm, and 40 nm chips rather than going back to the drawing board and investing in new technologies to compete with TSMC’s advanced 3 nm – 5 nm chips.
GlobalFoundries' headquarters in Malta, New York. Image (modified) courtesy of GlobalFoundries
GF has increased existing domestic manufacturing capacity in another way, too. The company has invested in gallium nitride (GaN) RF chips, a wide-band semiconductor technology that outperforms silicon in terms of thermal resistance and durability. GF's development of GaN devices at its Essex Junction, Vermont, facility would have been impossible without a $30 million government grant to shorten the time to market for the GaN RF technology.
A GF facility in upstate New York is also getting a boost—this time, from collaborating with Qualcomm. The U.S. mobile chipmaker has previously agreed to spend $4.2 billion on chips made by GF for Qualcomm's 5G transceivers, automotive products, and IoT connectivity.
I also posted not long ago about NASA and the 22nm FDSOI requirements plus that there wrote the same chip was often used for automotive.
Well, just seen this older solicitation with some further support for the 22nm. Ends mid this year.
Radiation Tolerant Standard Cell Library in a 22nm FDSOI CMOS Process
Award Information
Agency:National Aeronautics and Space Administration
Branch:N/A
Contract:80NSSC21C0516
Agency Tracking Number:205369
Amount:$759,984.00
Phase
Program:SBIR
Solicitation Topic Code:Z2
Solicitation Number:SBIR_20_P2
Timeline
Solicitation Year:2020
Award Year:2021
Award Start Date (Proposal Award Date):2021-07-27
Award End Date (Contract End Date):2023-07-26
Small Business Information
ALPHACORE INC
304 South Rockford Drive
Tempe, AZ 85281-0000
United States
NASA is seeking radiation tolerant standard cell libraries for processes below 28nm that are suitable for NASA missions in the natural space environment. As a response, Alphacore proposes to develop for NASA an innovative library of radiation-hardened (rad-hard) standard cells implemented in the GlobalFoundries (GF) 22nm fully depleted silicon on insulator (FDSOI) CMOS fabrication process (GF 22FDX).nbsp;Alphacore has been able to have the first version of the radiation tolerant library important cells already designed by the end of the Phase I program. The Phase I program provided an excellent basis for the Phase II program in which an extended, innovative radiation-tolerant standard cell library will be designed, fabricated, and tested. The library elements will be taped in the middle of the Phase II program and knowing the fabrication time for chips in the 22FDX program is 2.5 - 3 months, the evaluations can start approximately 3.5 months after the tapeout date. The elements will be tested for functionality, circuit performance and radiation hardness, including both single event effects (SEE) and total ionizing dose (TID).During Phase II, Alphacore will not only support CMOS IP development for the most harsh NASA mission environments (low temperatures and high radiation), but also covering critical immediate needs in other areas such as the High-Performance Spaceflight Computing (HPSC) Chiplet, a radiation-hardened multi-core processor. NASA also will be developing single board computer based on this chiplet. These components are supposed to bring NASA (and DoD) a two orders of magnitude improvement in their space computing capabilities.
I often go back to the Renesas tape out Ann to reread its context.
Key question for me is the following couple of paragraphs plus the common theme we are now seeing with 22nm both through Renesas in Dec and by BRN with GF...
They way I read the below is that Renesas have had someone come to them and request the tape out and device but open to interpretation I guess.
The other part is that Renesas will let the mkt decide on something like this as to the uptake and then decide if they will bring in house for control or just continue external 3rd parties to run with it.
Renesas tapes out spiking neural network chip
Renesas is taping out a chip using Brainchip's spiking neural network (SNN) technology and expanding internal and foundry capacity
“Now you have accelerators for driving AI with neural processing units rather than a dual core CPU. We are working with a third party taping out a device in December on 22nm CMOS,” said Chittipeddi.
The take up of the technology depends on the market adoption, he says.
“We want to see where the market reception is the highest, that is what determines whether we bring things in house or through a third party.”
Now...another recent article I was just reading is about GF and quite interesting group of names discussed including BRN.
Be nice to get in with Purdue as well. Intel in there and our own K Carlson published papers on STDP in 2011 via Purdue.
In a Series of Agreements, GlobalFoundries Buoys Chip Supply From Home
3 days ago by Biljana Ognenova
Tracking the goals of the CHIPS+ Act, GlobalFoundries is strengthening the U.S. semiconductor ecosystem with several partnerships and acquisitions.
The U.S. semiconductor manufacturing industry is beginning to reap the benefits of the CHIPS+ Act passed in July 2022. In a series of collaborations and acquisitions, GlobalFoundries (GF) is progressing toward one of the act's goals: to strengthen a U.S.-based semiconductor supply chain and reduce dependence on Asia-based manufacturers. Throughout the chip shortage, GF claims it has remained profitable by steadying productivity, containing cost, differentiating products, and doubling down on supply chain security.
A comparison of GF's and TSMC's EBIT margins. Image courtesy of Seeking Alpha
GlobalFoundries attributes its upward profitability to its strategic approach to chipmaking and its local partnership with key industry sectors, including automotive, memory, and computing as well as education.
GlobalFoundries Partners With GM, Renesas, and Purdue
In a major win for GF in the automotive sector, the foundry signed an agreement with GM on Feb. 9, 2023, to dedicate a manufacturing corridor at GF's upstate New York facility for GM’s key chip suppliers.
GF also secured more production capacity by acquiring Renesas' NVM resistive RAM technology. Specifically, GF now owns Renesas' proprietary Conductive Bridging Random Access Memory (CBRAM) technology, a low-power memory solution built for home and industrial IoT and mobile devices.
GF has secured wins in the education and R&D sector as well, partnering with Purdue University on a semiconductor education program. Purdue was among the first institutions to use the funds from the CHIPS+ Act to finalize a university-level semiconductor program. Now, with this new collaboration, Purdue's staff and students will use GlobalFoundries' facility and resources to create innovative, interdisciplinary solutions for advanced semiconductors and microelectronics. The partnership between GF and Purdue will provide next-generation professionals with hands-on expertise overseen at GF Lab.
GF Supports Next-gen Vision and Computing Technologies
GF is also championing the computing sector, namely with BrainChip's Akida neuromorphic chip built on 22 nm fully depleted silicon-on-insulator (FD-SOI) technology.
Akida architecture. Image courtesy of BrainChip
BrainChip is the world’s first company to develop ultra-low-power, event-based, neuromorphic AI IP to be used for always-on sensor applications. The AKD1500 chip was built on GF’s low-leakage FD SOI platform, promising an array of applications that don't overload the CPU.
Vision sensor specialist Oculi also recently announced a strategic partnership with GF, commissioning the foundry to manufacture its single-chip, intelligent software-defined vision sensor. The new sensor will be based on GF’s 55LPx, a platform that supports RF, ultra-low power, embedded NVM, and high-voltage BCDLite (a process technology).
Expanding U.S.-based Semiconductor Production
While GlobalFoundries relies on five-year agreements to ship chips from storage facilities in Dresden and Singapore, the foundry also plans to expand three U.S. locations, including one in Vermont and two in New York. Specifically, GF is broadening the scope of existing facilities to make 12 nm, 28 nm, and 40 nm chips rather than going back to the drawing board and investing in new technologies to compete with TSMC’s advanced 3 nm – 5 nm chips.
GlobalFoundries' headquarters in Malta, New York. Image (modified) courtesy of GlobalFoundries
GF has increased existing domestic manufacturing capacity in another way, too. The company has invested in gallium nitride (GaN) RF chips, a wide-band semiconductor technology that outperforms silicon in terms of thermal resistance and durability. GF's development of GaN devices at its Essex Junction, Vermont, facility would have been impossible without a $30 million government grant to shorten the time to market for the GaN RF technology.
A GF facility in upstate New York is also getting a boost—this time, from collaborating with Qualcomm. The U.S. mobile chipmaker has previously agreed to spend $4.2 billion on chips made by GF for Qualcomm's 5G transceivers, automotive products, and IoT connectivity.
I also posted not long ago about NASA and the 22nm FDSOI requirements plus that there wrote the same chip was often used for automotive.
Well, just seen this older solicitation with some further support for the 22nm. Ends mid this year.
Radiation Tolerant Standard Cell Library in a 22nm FDSOI CMOS Process
Award Information
Agency:National Aeronautics and Space Administration
Branch:N/A
Contract:80NSSC21C0516
Agency Tracking Number:205369
Amount:$759,984.00
Phase
Program:SBIR
Solicitation Topic Code:Z2
Solicitation Number:SBIR_20_P2
Timeline
Solicitation Year:2020
Award Year:2021
Award Start Date (Proposal Award Date):2021-07-27
Award End Date (Contract End Date):2023-07-26
Small Business Information
ALPHACORE INC
304 South Rockford Drive
Tempe, AZ 85281-0000
United States
NASA is seeking radiation tolerant standard cell libraries for processes below 28nm that are suitable for NASA missions in the natural space environment. As a response, Alphacore proposes to develop for NASA an innovative library of radiation-hardened (rad-hard) standard cells implemented in the GlobalFoundries (GF) 22nm fully depleted silicon on insulator (FDSOI) CMOS fabrication process (GF 22FDX).nbsp;Alphacore has been able to have the first version of the radiation tolerant library important cells already designed by the end of the Phase I program. The Phase I program provided an excellent basis for the Phase II program in which an extended, innovative radiation-tolerant standard cell library will be designed, fabricated, and tested. The library elements will be taped in the middle of the Phase II program and knowing the fabrication time for chips in the 22FDX program is 2.5 - 3 months, the evaluations can start approximately 3.5 months after the tapeout date. The elements will be tested for functionality, circuit performance and radiation hardness, including both single event effects (SEE) and total ionizing dose (TID).During Phase II, Alphacore will not only support CMOS IP development for the most harsh NASA mission environments (low temperatures and high radiation), but also covering critical immediate needs in other areas such as the High-Performance Spaceflight Computing (HPSC) Chiplet, a radiation-hardened multi-core processor. NASA also will be developing single board computer based on this chiplet. These components are supposed to bring NASA (and DoD) a two orders of magnitude improvement in their space computing capabilities.
"That said, I don't think we can expect Sean to pull rabbits out of the hat while they are still in their shells"
.. I agree.
Nice one TLS!
Aren’t we getting short changed with this Edge Impulse video? We are excellent at object detection and tracking, yet another company, Silabs amongst others is getting recognition for its system. Also Renesas. Or am I missing something here?Watch the Video! In the LinkedIn link.
Amir start talking about BrainChip toward the last two minutes.
View attachment 30515 View attachment 30516 View attachment 30517
Amir Sherman on LinkedIn: Checkout my Presentation at the Mext Semiconductors Event. The "new"…
Checkout my Presentation at the Mext Semiconductors Event. The "new" Semiconductors market with focus on EdgeML & EdgeAI and how Edge Impulse can help…www.linkedin.com
Learning 🏖
by AICorespot Editorial Team 
Hi Dhm,
Also don’t forget, edge impulse have referred to akida as science fiction, and we are the first and only “strategic” IP company.
www.edgeimpulse.com
Just a refresher for those who haven’t/didn’t watch these-Also don’t forget, edge impulse have referred to akida as science fiction, and we are the first and only “strategic” IP company.
“BrainChip the first strategic IP partner on the Edge Impulse platform”
Edge Impulse and BrainChip Partner to Further AI Development with Support for the Akida Platform
Edge Impulse has announced official support for BrainChip's neural processor AI IP.
www.youtube.com
Just a refresher for those who haven’t/didn’t watch these-
“New IP innovation erasing limits”
From 58:50 timeframe-
Edge Impulse Imagine 2022 Day 1
The edge ML conference 00:18:10 | CEO Keynote - Zach Shelby, Edge Impulse01:05:03 | Industry Welcome - Simon Segars01:13:23 | Application of Edge ML in the S...
And then Rob telson presenting that day-
To read something like this here in German almost brings tears to my eyes. Just - WoW
MENU Neuromorphic camera and machine learning aid nanoscopic imagingIn a new study, researchers at the Indian Institute of Science (IISc) show how a brain-inspired image sensor can go beyond the diffraction limit of light to detect miniscule objects such as cellular components or nanoparticles invisible to current microscopes. Their novel technique, which combines optical microscopy with a neuromorphic camera and machine learning algorithms, presents a major step forward in pinpointing objects smaller than 50 nanometers in size. The results are published in Nature Nanotechnology. Since the invention of optical microscopes, scientists have strived to surpass a barrier called the diffraction limit, which means that the microscope cannot distinguish between two objects if they are smaller than a certain size (typically 200-300 nanometers). Their efforts have largely focused on either modifying the molecules being imaged, or developing better illumination strategies – some of which led to the 2014 Nobel Prize in Chemistry. “But very few have actually tried to use the detector itself to try and surpass this detection limit,” says Deepak Nair, Associate Professor at the Centre for Neuroscience (CNS), IISc, and corresponding author of the study. 
Transformation of cumulative probability density of ON and OFF processes allows localisation below the limit of classical single particle detection (Credit: Mangalwedhekar et al, 2023) Measuring roughly 40 mm (height) by 60 mm (width) by 25 mm (diameter), and weighing about 100 grams, the neuromorphic camera used in the study mimics the way the human retina converts light into electrical impulses, and has several advantages over conventional cameras. In a typical camera, each pixel captures the intensity of light falling on it for the entire exposure time that the camera focuses on the object, and all these pixels are pooled together to reconstruct an image of the object. In neuromorphic cameras, each pixel operates independently and asynchronously, generating events or spikes only when there is a change in the intensity of light falling on that pixel. This generates sparse and lower amount of data compared to traditional cameras, which capture every pixel value at a fixed rate, regardless of whether there is any change in the scene. This functioning of a neuromorphic camera is similar to how the human retina works, and allows the camera to “sample” the environment with much higher temporal resolution – because it is not limited by a frame rate like normal cameras – and also perform background suppression. “Such neuromorphic cameras have a very high dynamic range (>120 dB), which means that you can go from a very low-light environment to very high-light conditions. The combination of the asynchronous nature, high dynamic range, sparse data, and high temporal resolution of neuromorphic cameras make them well-suited for use in neuromorphic microscopy,” explains Chetan Singh Thakur, Assistant Professor at the Department of Electronic Systems Engineering (DESE), IISc, and co-author. 
View of the microscopy setup (Credit: Rohit Mangalwedhekar) In the current study, the group used their neuromorphic camera to pinpoint individual fluorescent beads smaller than the limit of diffraction, by shining laser pulses at both high and low intensities, and measuring the variation in the fluorescence levels. As the intensity increases, the camera captures the signal as an “ON” event, while an “OFF” event is reported when the light intensity decreases. The data from these events were pooled together to reconstruct frames. To accurately locate the fluorescent particles within the frames, the team used two methods. The first was a deep learning algorithm, trained on about one and a half million image simulations that closely represented the experimental data, to predict where the centroid of the object could be, explains Rohit Mangalwedhekar, former research intern at CNS and first author of the study. A wavelet segmentation algorithm was also used to determine the centroids of the particles separately for the ON and the OFF events. Combining the predictions from both allowed the team to zero in on the object’s precise location with greater accuracy than existing techniques. “In biological processes like self-organisation, you have molecules that are alternating between random or directed movement, or that are immobilised,” explains Nair. “Therefore, you need to have the ability to locate the centre of this molecule with the highest precision possible so that we can understand the thumb rules that allow the self-organisation.” The team was able to closely track the movement of a fluorescent bead moving freely in an aqueous solution using this technique. This approach can, therefore, have widespread applications in precisely tracking and understanding stochastic processes in biology, chemistry and physics. REFERENCE: Mangalwedhekar R, Singh N, Thakur CS, Seelamantula CS, Jose M, Nair D, Achieving nanoscale precision using neuromorphic localization microscopy, Nature Nanotechnology (2023). |
GO BRN 
