Hi Bravo,Maybe some future competition on the horizon? @Diogense might be able to explain how analog neuromorphic computing compares? It says here that the " project, which started last month, aims to develop an analog neuromorphic computing approach based on oscillatory neural networks (ONNs) that seamlessly interfaces with sensors and processes analog data without any analog-to-digital conversion."
I remember that @TECH mentioned Simon Thorpe was working on a project with ONN's but I don't know if it's this particular one.
![]()
Oscillating neural networks for low power analog edge AI
Technology News | February 6, 2023
By Nick Flaherty
MATERIALS & PROCESSES AI
A European project including IBM and BMW is developing a new type of analog neural network with phase change materials to reduce the power consumption of machine learning at the edge of the network.
The new technique couples oscillating neural networks (ONN) with phase change materials and could reduce power consumption by a factor of 100 to 1000 say the researchers in the PHASTRAC (Phase Change Materials for Energy Efficient Edge Computing) project.
The project, which started last month, aims to develop an analog neuromorphic computing approach based on oscillatory neural networks (ONNs) that seamlessly interfaces with sensors and processes analog data without any analog-to-digital conversion.
- In-memory neuromorphic AI slashes power consumption
- imec, Leti form European hub for AI
- ST hints at analog in-memory computing chip
The oscillating neurons will be implemented with vanadium dioxide (VO2) phase change material coupled with synapses implemented with bilayer resistive RRAM memories using molybdenum and hafnium dioxide (Mo/HfO2).
The project aims to develop new devices for implementing the ONN architecture and processing the analog sensor data. It is led by researchers from Technical University in Eindhoven, who held a session on ONN technology at the European Conference on High-performance Embedded Architecture and Compilation (HiPEAC) Conference last month in Toulouse. They are working with researchers from BMW, IBM Research in Zurich and the Pázmány Péter Catholic University in Budapest.
![]()
Oscillating neural networks for low power analog edge AI
A European project including IBM and BMW is developing a new type of analog neural network with phase change materials for edge AIwww.eenewseurope.com
Hey, what did I do?Hi Bravo,
Apologies for that German chap.
That project looks to me to be more in the "R" phase of R&D.
"The project aims to develop new devices for implementing the ONN architecture and processing the analog sensor data. It is led by researchers from Technical University in Eindhoven"
It is a university project and universities seem to br enthralled by the close analogy between wetware and analog neurons. They still have to come to grips with the inherent variability of ReRAM/MemRistors.
Not commercial whthin 5 years +.
Thanks Diogenese, Our 5 year advantage remains!Hi Bravo,
Apologies for that German chap.
That project looks to me to be more in the "R" phase of R&D.
"The project aims to develop new devices for implementing the ONN architecture and processing the analog sensor data. It is led by researchers from Technical University in Eindhoven"
It is a university project and universities seem to br enthralled by the close analogy between wetware and analog neurons. They still have to come to grips with the inherent variability of ReRAM/MemRistors.
Not commercial whthin 5 years +.
I have been thinking about Tony Dawe's letter.This is a letter from TD published by @YngInvstr in response to his/her question on the crapper.
View attachment 28894
Relax Pal. People are just talking and expressing themselves. Who said it has to be just all one’s and zero’s or dots which may or may not be relevant to anything? Sometimes some of us get a little carried away, but that too is just humans being human. When we are all just androids on steroids it gets pretty boring. Variety is the spice of life.Wow what a thread this has become.
Do people need reminding that if you want to talk about exceeding your risk tolerance, go HERE.
Or if you feel like your self made timelines are not being kept by a company who didn't promise you anything, try going HERE or HERE.
Edit: the final "here" was supposed to be the crapper, how sad the links are blocked.
Thanks Dio, much appreciated. As always your input is welcomed and extremely valuable to those of us who are not overly technically oriented. Thank you for being so generous with your responses.Hi Bravo,
Apologies for that German chap.
That project looks to me to be more in the "R" phase of R&D.
"The project aims to develop new devices for implementing the ONN architecture and processing the analog sensor data. It is led by researchers from Technical University in Eindhoven"
It is a university project and universities seem to br enthralled by the close analogy between wetware and analog neurons. They still have to come to grips with the inherent variability of ReRAM/MemRistors.
Not commercial whthin 5 years +.
I think the Company is well aware that pressure is building and will adapt its strategy appropriately to facilitate revenue as rapidly as possible.I have been thinking about Tony Dawe's letter.
I thought that Akida2000 was the next chip to be taped out, however in TD's letter he states that the Akida1500 tape out was because of direct engagement with customers. I can't imagine BRN would have taped out Akida1500 without having firm commitments from licensee's or customers.
This raises a few questions.
I believe we have a good chance that an announcement will come very soon after Akida1500 has passed muster, and I think it will be the MOA announcements.
- Who could want the Akida1500 so badly that BRN moved it ahead of the Akida2000 tape out.
- Once the Akida1500 has been examined, tested and passed, should we expect a new License Agreement? or
- Can we expect actual orders for the Akida1500 in the very near future, because the chip was MADE TO ORDER for someone?
I have been thinking about Tony Dawe's letter.
I thought that Akida2000 was the next chip to be taped out, however in TD's letter he states that the Akida1500 tape out was because of direct engagement with customers. I can't imagine BRN would have taped out Akida1500 without having firm commitments from licensee's or customers.
This raises a few questions.
I believe we have a good chance that an announcement will come very soon after Akida1500 has passed muster, and I think it will be the MOA announcements.
- Who could want the Akida1500 so badly that BRN moved it ahead of the Akida2000 tape out.
- Once the Akida1500 has been examined, tested and passed, should we expect a new License Agreement? or
- Can we expect actual orders for the Akida1500 in the very near future, because the chip was MADE TO ORDER for someone?
Also mega chips did back end design on akd1500also remember megachips report the other day - aiming to be LAUNCHING PRODUCTS - note the plural
they are also projected to increase revenue from 56000 M Yen to 72000 M Yen which - if my maths is right, about 100M USD increase in sales
View attachment 28913
Thanks for trying Damo. They will just scroll past and keep repeating themselves despite so many well thought out and reasoned arguments as to why they don’t need to.Wow what a thread this has become.
Do people need reminding that if you want to talk about exceeding your risk tolerance, go HERE.
Or if you feel like your self made timelines are not being kept by a company who didn't promise you anything, try going HERE or HERE.
Edit: the final "here" was supposed to be the crapper, how sad the links are blocked.
Defititely agree. No one is going to all the effort to design, plan and tapeout based on a casual remark by a potential customer! it would be from some solid engagement and mutual development that requires it on the FDX platform. $$$$I have been thinking about Tony Dawe's letter.
I thought that Akida2000 was the next chip to be taped out, however in TD's letter he states that the Akida1500 tape out was because of direct engagement with customers. I can't imagine BRN would have taped out Akida1500 without having firm commitments from licensee's or customers.
This raises a few questions.
I believe we have a good chance that an announcement will come very soon after Akida1500 has passed muster, and I think it will be the MOA announcements.
- Who could want the Akida1500 so badly that BRN moved it ahead of the Akida2000 tape out.
- Once the Akida1500 has been examined, tested and passed, should we expect a new License Agreement? or
- Can we expect actual orders for the Akida1500 in the very near future, because the chip was MADE TO ORDER for someone?
also remember megachips report the other day - aiming to be LAUNCHING PRODUCTS - note the plural
they are also projected to increase revenue from 56000 M Yen to 72000 M Yen which - if my maths is right, about 100M USD increase in sales
View attachment 28913
The thing that sent me into a head spin was TD's statement - "the Akida1500 tape out was because of direct engagement with customers".I don't think so VG;
1500 was meant to come ahead of 2000.
And note how AKD2000 is the ''optimized version'' of AKD1500.
View attachment 28915
I don't think theres any straws involved. Remember what happened with the first iteration of the product after it was given to EAP's???The thing that sent me into a head spin was TD's statement - "the Akida1500 tape out was because of direct engagement with customers".
Why would TD say that if it had nothing to do with customer engagement as it was the next scheduled role out anyway.
I'm probably clutching at straws.
EAPs would have the option imo.The thing that sent me into a head spin was TD's statement - "the Akida1500 tape out was because of direct engagement with customers".
Why would TD say that if it had nothing to do with customer engagement as it was the next scheduled role out anyway.
I'm probably clutching at straws.