BRN Discussion Ongoing

Rskiff

Regular
  • Haha
  • Like
  • Fire
Reactions: 14 users

M_C

Founding Member
  • Like
  • Haha
  • Thinking
Reactions: 6 users

Boab

I wish I could paint like Vincent
  • Like
  • Haha
Reactions: 6 users

BaconLover

Founding Member


We have read this a few months ago, but a refresher;

The 5-nm semiconductor applied with the extreme ultraviolet (EUV) process is a high-tech product that only a small number of companies, such as Samsung Electronics and Taiwan Semiconductor Manufacturing Company (TSMC), can produce.
These will be used in the Tesla infotainment (IVI) product family, which utilises semiconductors such as processors, neural network processing units (NPUs), security integrated circuits, memories, and display driving chips (DDI).
 
Last edited:
  • Like
  • Fire
  • Love
Reactions: 23 users

Slymeat

Move on, nothing to see.
Another true story.

Where I went to school we had quite a few PNG students who had come to Australia on Mobil Oil scholarships. When I finished school I travelled around PNG with one of the native students (Philip), staying in the towns and villages of fellow students. We even hitch hiked up the Highlands Highway and back. Finally we reached Phillip's village of Paruai in New Ireland where I spent some time and became involved in normal activities including gathering Sago from the wild sago palm. It would be stored wrapped in banana leaves and often cooked with the evening meal.

After a while the sago would get weevils in it but, but when cooked up it went down just as well, providing a little extra protein.

When i returned to Australia I found myself at the Queensland Agricultural College, and at one point was living in a caravan in the yard of the Forest Hill Produce Agent who I helped out with yard management and played tennis with them on weekends for the Forest Hill tennis club. The home team provided afternoon tea.

Living on my own in the caravan my muesli lasted a long time ... and eventually got weevils in it. Recalling my time in PNG I continued to eat it until I could eventually taste the weevils. Then I started to cook it up like porridge, but I could eventually even taste the weevils in the 'porridge'. So I made ANZAC biscuits out of it and took it to tennis. They were a big hit with multiple people wanting to know my recipe. Mind you I tucked into the jam and scones ... and the recipe remained my secret.

Deena
So what I’m hearing here is that I wasted a lot of time hand plucking the weevils out of oats my father bought “as a bargain”. I could have simply made porridge out of them as is, and get some added protein to further add to the bargain.

Maybe my father picked up a lack of aversion to weevil-infested food in his time spent on the naval base at Manus Island (PNG) after WWII.🤔

Until now, I thought my humble beginnings were quite unique, especially on investing forums.
 
  • Like
  • Love
  • Haha
Reactions: 15 users

jk6199

Regular
So what I’m hearing here is that I wasted a lot of time hand plucking the weevils out of oats my father bought “as a bargain”. I could have simply made porridge out of them as is, and get some added protein to further add to the bargain.

Maybe my father picked up a lack of aversion to weevil-infested food in his time spent on the naval base at Manus Island (PNG) after WWII.🤔

Until now, I thought my humble beginnings were quite unique, especially on investing forums.
Probably explains why so many are only too willing to be so generous with information. Not too many born with silver spoons in their mouths.
 
  • Like
  • Love
Reactions: 17 users
Sorry for letting my legal training get in the way of @MC🐠 and his speculation but there is some significance in the fact that Brainchip has adopted and republished on its social media accounts the following extract from an article by a representative of Renesas without attribution other than the link at the end.

I won’t bore you with the legal mumbo jumbo but it is significant and we can take it that Brainchip is saying this is all about us including the first paragraph:

BrainChip

BrainChip

4 天前

The other side is the embedded AI with DRP dynamically reconfigurable processor for vision solutions. That’s a feedforward neural network rather than a convolutional neural network (CNN), and it offers reasonable 0.5TOPS to 10TOPS at very low power compared to day the Nvidia or Intel equivalent.

From my perspective RISC-V will evolve into that areas in the not too distant future. “So that’s kind of the plan that we have, to evolve that into that landscape,”

he said. “At the very low end, we have added an ARM M33 MCU and spiking neural network with BrainChip core licensed for selected applications

– we have licensed what we need to license from BrainChip including the software to get the ball rolling.”

My opinion only DYOR
FF

AKIDA BALLISTA
 
  • Like
  • Fire
  • Love
Reactions: 54 users

Slymeat

Move on, nothing to see.
This link is very interesting particularly the two recently opened new Institutional Investors:


My opinion only DYOR
FF

AKIDA BALLISTA
What a great find, thanks @Fact Finder .

I love the following line, it reads institutional investors have faith. I am a little concerned one institution is a small cap fund. They may not be able to hold a position for too much longer. But then the large cap, and blue chip, funds will take over from them soon.

Institutional Owners15 total, 15 long only, 0 short only, 0 long/short
 
  • Like
  • Fire
  • Love
Reactions: 13 users

Boab

I wish I could paint like Vincent


We have read this a few months ago, but a refresher;

The 5-nm semiconductor applied with the extreme ultraviolet (EUV) process is a high-tech product that only a small number of companies, such as Samsung Electronics and Taiwan Semiconductor Manufacturing Company (TSMC), can produce.
These will be used in the Tesla infotainment (IVI) product family, which utilises semiconductors such as processors, neural network processing units (NPUs), security integrated circuits, memories, and display driving chips (DDI).
Interesting how Doug Fairbairn pointed out(in the megachips podcast posted this morning) that 5-nm is so much more expensive and it's use is limited ( probably because of its cost?)
 
  • Like
  • Love
  • Fire
Reactions: 13 users

Slymeat

Move on, nothing to see.
  • Like
  • Fire
  • Love
Reactions: 13 users

Murphy

Life is not a dress rehearsal!
Interesting how Doug Fairbairn pointed out(in the megachips podcast posted this morning) that 5-nm is so much more expensive and it's use is limited ( probably because of its cost?)
EXPONENTIALLY more expensive! Even moving down from 22nm to 16nm was way, way more expensive. On another note, just subscribed for a year with Zeebot TSE, and this post about the Podcast justifies the yearly subscription on its own!! Get on board for this cornucopia of shared info!!

If you don't have dreams, you can't have dreams come true!
 
  • Like
  • Love
  • Fire
Reactions: 23 users

BaconLover

Founding Member
Interesting how Doug Fairbairn pointed out(in the megachips podcast posted this morning) that 5-nm is so much more expensive and it's use is limited ( probably because of its cost?)
tsmc-5nm.jpg


Apologies, I do not know the source, was in my downloads.
 
Last edited:
  • Like
  • Love
  • Fire
Reactions: 20 users
Interesting how Doug Fairbairn pointed out(in the megachips podcast posted this morning) that 5-nm is so much more expensive and it's use is limited ( probably because of its cost?)
Do you remember a post I put up recently which was a reply from Peter van der Made addressing whether one of the many pretenders in this space was actually real competition.

If so you will recall he mentioned what the former CEO Mr. Dinardo discussed in one of his podcasts in 2020 that AKIDA’s performance would keep increasing as you scaled down from 28nm to 14nm to 7nm etc;

As I remember it Peter van der Made referenced a 100 or 1,000 times improvement. I am sure someone will throw it up here but the exact numbers are not really important.

What Doug in my opinion is referencing is that with the solutions they have via AKIDA you can get all the performance you need and more at 22nm without the huge cost of 5nm.

The primary reason companies are chasing these small chips is to try and outrun Moore’s Law but at what cost to their competitive advantage.

Those companies that are secretly developing their products with AKIDA IP at 22nm and above will be able to crush their competitors on price because of the performance they can offer at far less cost simply because without AKIDA competitors have to go with 7nm or less to stay in the race.

The other factor is that chip yields fall off the lower down they go and as I posted the other day Samsung has had so many problems at 4nm they have gone to TSMC.

Going back to 2020 Intel was talking about giving up on the foundry side of things because they just could not achieve sufficient yields at 7nm and 14nm to remain competitive.

I think it is true to say as someone did earlier today that the full commercial advantage offered by Brainchip is not generally understood.

My opinion only DYOR
FF

AKIDA BALLISTA
 
  • Like
  • Love
  • Fire
Reactions: 61 users

Labsy

Regular
This is an interesting read about autonomous drones. The interest and funding in Brainchip from Defence is obvious and the current situation has them all scrambling to implement the tech.

Right time, right place!

TECH & SCIENCE

China’s drone carrier hints at ‘swarm’ ambitions for Pacific​

By
AFP
Published
June 8, 2022

Last month, Chinese researchers published a drone swarm experiment allegedly showing devices autonomously navigating a dense patch of bamboo forest

Last month, Chinese researchers published a drone swarm experiment allegedly showing devices autonomously navigating a dense patch of bamboo forest - Copyright AFP/File Lillian SUWANRUMPHA
Didier LAURAS
Officially it is just a research vessel, but China’s newly unveiled drone carrier is a clear sign Beijing is rushing to deploy an autonomous swarm of unmanned devices in its push for military supremacy in the Pacific Ocean.
State media last month showed the launching of the Zhu Hai Yun — “Zhu Hai Cloud” — capable of transporting an unspecified number of flying drones as well as surface and submarine craft, and operating autonomously thanks to artificial intelligence.
The 89-metre (292-foot) ship would be operational by year-end with a top speed of 18 knots, vastly increasing China’s surveillance potential of the vast Pacific area it considers its zone of influence.
“The vessel is not only an unprecedented precision tool at the frontier of marine science, but also a platform for marine disaster prevention and mitigation, seabed precision mapping, marine environment monitoring, and maritime search and rescue,” Chen Dake, lab director at the firm that built the carrier, told China Daily.
Armies worldwide see drone squadrons as key players in combat, able to overwhelm defence systems by sheer numbers and without putting soldiers’ lives at risk, such as with more expensive jets or tanks.
“It’s probably a first-of-its-kind development but other navies across the world, including the US Navy, are experimenting with remote warfare capabilities in the maritime domain,” said US Army Lieutenant Colonel Paul Lushenko, who is also an international relations specialist at Cornell University in New York.
Even if the vessel’s actual capabilities remain to be seen, Beijing is broadcasting its intent to cement territorial claims in the region, as seen with the security partnership agreed last month with the Solomon Islands northeast of Australia.
“It’s definitely imposing, provocative, escalatory and aggressive,” Lushenko told AFP.
– Collective intelligence –
Building fleets of autonomous and relatively inexpensive drones would greatly augment China’s ability to enforce so-called anti-access and area denial (A2-AD) in the Pacific, with the aim of weakening decades of US influence.
Unlike traditional aircraft carriers or destroyers carrying hundreds of troops, the drone carrier could itself navigate for longer periods while sending out devices that create a surveillance “net,” potentially able to fire missiles as well.
The Zhu Hai Yun could also improve China’s mapping of the seafloor, providing a covert advantage for its submarines.
“These are capabilities that are likely to be critical in any future conflicts that China wages, including over the island of Taiwan,” strategists Joseph Trevithick and Oliver Parken wrote on the influential War Zone site.
Beijing has made no secret of its desire to wrest control of Taiwan, and military experts say it is closely watching the West’s response to the Russian invasion of Ukraine to gauge how and when it might make its move.
And last month, Chinese researchers published a drone swarm experiment allegedly showing 10 devices autonomously navigating a dense patch of bamboo forest, without crashing into the trees or each other.
“The ultimate goal is something that has a collective intelligence,” said Jean-Marc Rickli, head of risks at the Geneva Centre for Security Policy.
“The analogy is a bit like a school of fish. They create forms in the water that are not the decision of any single fish, but the result of their collective intelligence,” he told AFP.
– Game-changer –
It would be a big technological advance from current weapons, which can be programmed and semi-autonomous but must have human operators to react to unexpected challenges.
A fleet of self-navigating drones could in theory incapacitate defence systems or advancing forces by sheer numbers, saturating combat zones on land or at sea until an opponent’s arsenal is depleted.
“A conventional attack becomes impossible when you’re facing dozens, hundreds or thousands of devices that are much cheaper to develop and operate than heavy conventional weapons,” Rickli said.
Noting this profound shift in modern warfare, a RAND Corporation study from 2020 found that while unmanned vehicles need significant improvements in onboard processing, “the overall computing capability required will be modest by modern standards — certainly less than that of a contemporary smartphone.”
“A squadron of approximately 900 personnel, properly equipped and trained, could launch and recover 300 L-CAATs every six hours, for a total of 1,200 sorties per day,” it said, referring to low-cost attributable aircraft technology — meaning devices so cheap an army can afford to lose them.
“We do have indications that China is making rapid capabilities development,” Lushenko said of Beijing’s new drone carrier.
“What we lack is empirical data to suggest that China’s one-party state can actually employ the ship in an integrated fashion in conflict.”


Read more: https://www.digitaljournal.com/tech...m-ambitions-for-pacific/article#ixzz7VwnfW34X
Just a left field question for all the tech guys out there... would a rad hardened akida driven SOC be affected by an EMP blast? If not, then could this be a strategic advantage in a future tech driven AI war? Hmmm 🤔🤫
 
  • Like
  • Fire
  • Thinking
Reactions: 8 users

Boab

I wish I could paint like Vincent
Do you remember a post I put up recently which was a reply from Peter van der Made addressing whether one of the many pretenders in this space was actually real competition.

If so you will recall he mentioned what the former CEO Mr. Dinardo discussed in one of his podcasts in 2020 that AKIDA’s performance would keep increasing as you scaled down from 28nm to 14nm to 7nm etc;

As I remember it Peter van der Made referenced a 100 or 1,000 times improvement. I am sure someone will throw it up here but the exact numbers are not really important.

What Doug in my opinion is referencing is that with the solutions they have via AKIDA you can get all the performance you need and more at 22nm without the huge cost of 5nm.

The primary reason companies are chasing these small chips is to try and outrun Moore’s Law but at what cost to their competitive advantage.

Those companies that are secretly developing their products with AKIDA IP at 22nm and above will be able to crush their competitors on price because of the performance they can offer at far less cost simply because without AKIDA competitors have to go with 7nm or less to stay in the race.

The other factor is that chip yields fall off the lower down they go and as I posted the other day Samsung has had so many problems at 4nm they have gone to TSMC.

Going back to 2020 Intel was talking about giving up on the foundry side of things because they just could not achieve sufficient yields at 7nm and 14nm to remain competitive.

I think it is true to say as someone did earlier today that the full commercial advantage offered by Brainchip is not generally understood.

My opinion only DYOR
FF

AKIDA BALLISTA
"Those companies that are secretly developing their products with AKIDA IP at 22nm and above will be able to crush their competitors on price because of the performance they can offer at far less cost simply because without AKIDA competitors have to go with 7nm or less to stay in the race."
What a fabulous sentence this is. Thank you FF
 
  • Like
  • Fire
  • Love
Reactions: 21 users

M_C

Founding Member
  • Like
  • Fire
  • Love
Reactions: 46 users

Slymeat

Move on, nothing to see.
View attachment 9155

Apologies, I do not know the source, was in my downloads.
Thanks for sharing that @BaconLover. Very telling indeed. I actually thought the price wasn’t quite so linear, with a sweet spot around 16-22 nm where many economies of the technology come into play. But this chart is representative enough for comparison purposes.

I don’t think per-wafer cost is the most important element. Using the supplied chart for guidance, price per transistor/processing element (for want of another word to imply what the size actually measures) is comparable for all sub 20 nm technologies, and actually is close to 1/10th the cost for 5 nm when compared with 40 nm, as 5 nm has 64 times the stuff in the same area, but cost is only about 6-7 times per wafer.

I believe a huge cost concern is with what you can embed your technology, as that is where huge cost savings result from marrying like with like. That is one of the reasons I am so happy that Weebit Nano has its ReRAM working in 28 nm and 22 nm as so much already exists in that technology, facilitating SOC manufacturing in pretty much a single process with no fab re-tooling.
 
  • Like
  • Love
Reactions: 13 users
''MegaChips have relationships with TSMC and Global Foundries.
We also work with ?UMC and SAMSUNG. We also worked with Intel but they are a bit down road the road for this kind of business''
Big take away for me, is that MegaChips, spearhead into the US markets, is based primarily on Edge A.I.
Brainchip and Quadric, are the two Edge A.I. technologies (which are complimentary, basically best used together) that they have chosen, for this purpose.

That is the magnitude of importantance and opportunity, that MegaChips sees, in the Edge A.I. Market.

I was really interested, to see if MegaChips had revealed another customer, in Samsung (Nintendo being the only one, the World knows about).
But MegaChips in this instance, is just a customer of Samsung chip foundries, so we are none the wiser (Samsung, is one of the chip foundries, that MegaChips, arranges for its customers chips to be made at).

Love the speed up feature on the podcast, these guys talk too slow.. 😛
 
Last edited:
  • Like
  • Fire
  • Love
Reactions: 26 users

Kachoo

Regular
AAnnnnnddddd .....................not a one off it seems......................

View attachment 9138
I believe that Anil's daughter works with Tesla. I could be wrong but I remember a post showing that a few years back. If this is false then ill ask to delete the post but i read this someplace im sure of it.
 
  • Like
  • Fire
  • Love
Reactions: 13 users
Top Bottom