Makeme 2020
Regular
Free to all Shareholders............lol"Cash receipts for a single qtr, especially when there is potentially missing spondoola, should not cause a 15% drop"..... it didnt, it caused a 21% drop.... WTFlol
BRN Discussion Ongoing
- Thread starter TechGirl
- Start date
Proga
Regular
I don't know mate. My thinking is the last paragraph in the CEO Statement - "Extended evaluations, decreased budgets, delayed introduction of new technology, these conditions have created a headwind" did the damage. Lumpy sales was a known.
I also think this has created an opportunity for those companies pushing ahead to get earlier fab slots.
Fact Finder
Top 20
I was not questioning your motives simply bringing my post up to date.
Sorry if this upset you.
I actually thought my post stood the test of time quite well and supported how well the company keeps to its time lines.
Obviously you have a different view and that is your right.
My right is to not agree with your view about the company’s performance.
As I always say my opinion only DYOR
FF
AKIDA BALLISTA
TheFunkMachine
seeds have the potential to become trees.
I haven't even bothered reading on this page today as I am pretty sure of the rubbish I will find. It’s a hard blow to swallow, but if you don’t have the conviction or guts to hold trough these dips you need to stop calling yourself a investor and start calling yourself a trader. Nothing wrong with either, just figure out what you are so that you can stop acting out of emotion 
I was expecting more then this in the 4c so it definitely was a little disappointing, but there is yet another 4c before 2023 and still room for preforming based on the statements explosive growth and second half of 2022.
Anyways. I have ridden these waves with Brainchip a few times now and it has always come back stronger. The last time the RSI was this low on the weekly we had our massive run next. It is now back at the over all upward trend line as depicted in my chart and it will be interesting to see how that plays out.
I am annoyed and frustrated to see my money disappearing in front of me, but I also am as convicted now about the prospects of BRN as I was 2 days ago so I will just have to be patient.
BTW I don’t know the last time I have actually been right with my charts so take this with a grain of salt. Much love
I was expecting more then this in the 4c so it definitely was a little disappointing, but there is yet another 4c before 2023 and still room for preforming based on the statements explosive growth and second half of 2022.
Anyways. I have ridden these waves with Brainchip a few times now and it has always come back stronger. The last time the RSI was this low on the weekly we had our massive run next. It is now back at the over all upward trend line as depicted in my chart and it will be interesting to see how that plays out.
I am annoyed and frustrated to see my money disappearing in front of me, but I also am as convicted now about the prospects of BRN as I was 2 days ago so I will just have to be patient.
BTW I don’t know the last time I have actually been right with my charts so take this with a grain of salt. Much love
Attachments
Xhosa12345
Regular
The revenue could be plenty later on in the quarter but not recieved yed. The 4 c this is purely cash. My concern was 2.5 to 3.5 in debtors (at least 2 .5) was not collected, and hence my email yesterday to the co, when the half yearly stated 90 day max expected terms, ie most of it shouldve been collected. But as i acknowledged to tony in my email he probably has a lot of emails to deal with so im not expecting a response in the short term. When i get one i will let u all know. I think that part deserves an explanation and theres no way that can breach a fkn nda... how hard to say we have not collected x dollars by 30 sept, but we have subsequently received it. Or if it was longer than 90 day terms, disclose that in the half yearly. That is what is pissing me off and must be addressed. As i stated previously i wish all companies were forced to release quarterly accrual accounts along with the 4cs, but never going to happen.
jtardif999
Regular
With respect what you are saying is a bit of rubbish. Do you think somehow they are not getting the word out about AKIDA technology, not through all of the worlds leading edge seminars, exhibitions and conferences - of which mostly BRN have pretty much had front-row seats and, not through establishing partnerships with the likes of ARM, Edge Impulse, SiFive and Prophesee?? Do you think that we should have more of a sales pipeline by now, indeed do you think the company should be seeing more actual sales now? I guess most here were probably hoping to see more of something show up in this 4c but realistically that something would only be some more licence fees and maybe a smattering of royalties beginning to trickle through? We could hope for royalties but the reality is they are still probably 2 quarters away from showing as realised cash on the books. Let’s look at this objectively, BRN have the best edge tech and the patents to lock in differentiation for the foreseeable future - we just need patience now and royalties will start to trickle before they flow, before eventually they become a raging torrent. The licence fees we expected to see in this 4c as specified as revenue in the half yearly will probably show in the next 4c (but it’s not a given). We all hope to see more licence fees added to that and I think we will. Stay confident and eventually you will be rewarded imo.
FabricatedLunacy
Regular
I’m a brnliever!!
But wow… Fridays used to be my most loved day. Then I started investing in BRN.
But wow… Fridays used to be my most loved day. Then I started investing in BRN.
I was not upset FF and I respect your view.
Everyone one of us have different opinions and views but at the end of the day want this company to do well.
Cheers.
Last edited:
WheresTheMonkey
Regular
Noooooooooooooooo.......................it does not mean no Revenue. It's a Cash Statement that's why.
BRN could have sold the Sydney Harbour Bridge and the Opera House to someone who has not yet paid the two invoices and it still won't show up in the Quarterly.
Yes it's that simple
SERA2g
Founding Member
My three biggest investments are on that list.Hey, Cheer up folks, we're NUMBER ONE in 2 catagories
View attachment 20436
We're now in the same price range as this time last year.
Nothing has happened since then, to add value to Company's prospects huh?..
Shorters are revelling now, the Fools all have boners.
Today is their day, no doubt about that..
Onwards ho!
You’ll see them in my signature. Rough day haha.
Fact Finder
Top 20
Once again I read the news and knew these things were occurring in the semiconductor space and have posted for months about how even though Brainchip was playing in this industry it had the advantage of:
1. Not carrying any debt;
2. Not carrying inventory because it sells IP that fits in a filing cabinet or on a memory stick;
3. It has cash and guaranteed further cash from LDA Capital and an offer to further extend the agreement if Brainchip wants to do so;
4. A product in AKIDA IP which can be added to existing semiconductors of the customer dramatically improving performance and energy consumption;
5. A product in the form of IP which can be priced to suit market conditions as it in essence has no ongoing cost of production;
6. A market place where automotive and other industries are being required to cut energy use by 2030 under legislative mandates which make AKIDA IP attractive.
So if it was the CEO Sean Hehir’s statement that
"Extended evaluations, decreased budgets, delayed introduction of new technology, these conditions have created a headwind"
which caused todays collapse how is it that I alone in all the world already knew and posted about it.
I must live in an echo chamber and receive messages that no one else does from the Fourth Estate.
Once again all I can say is that those who did not already know these things should actually do their own research.
Perhaps as pompous as it sounds even read my posts.
What the CEO Sean Hehir said was old news publicly available to every shareholder for many months.
My opinion only DYOR
FF
AKIDA BALLISTA
WheresTheMonkey
Regular
Yep I use esuper as well. One simple checklist each year and it's all done for you for less than $1K
cosors
👀
To all the disappointed investors who have been crying here because after a year of rolling out a revolutionary new technology on the world market, there are still no continuously rising revenues in the millions: Are you now happy to have made the shorters rich? Such a lament here yesterday...
Who has benefited from it? Thanks to the many disappointed who did not get what they wanted, namely product development cycles of less than a year, we are now roughly back where I started. As if nothing had happened in the last year. Am I supposed to thank you for that, you faithful technology understanders? I am new to this sector. I have an idea of how long it takes to bring a product to market and also how long it takes for something completely new. And this is new to the world. But now we know who is just spouting big pithy words and shitting their trousers in fear. We will conquer the world's chip production in one year...
A deadly serious tip from me is to buy some Metaverse, which also wants to conquer the world. Go over there so I don't have to read any more of this whining.
Shorter's cheerleaders.
Yes, but...
Idiots
Who has benefited from it? Thanks to the many disappointed who did not get what they wanted, namely product development cycles of less than a year, we are now roughly back where I started. As if nothing had happened in the last year. Am I supposed to thank you for that, you faithful technology understanders? I am new to this sector. I have an idea of how long it takes to bring a product to market and also how long it takes for something completely new. And this is new to the world. But now we know who is just spouting big pithy words and shitting their trousers in fear. We will conquer the world's chip production in one year...
A deadly serious tip from me is to buy some Metaverse, which also wants to conquer the world. Go over there so I don't have to read any more of this whining.
Shorter's cheerleaders.
Yes, but...
Idiots
Painful day. Just no buyers today because of the bad 4c. When you are getting low ball offers on your house because of market conditions do you sell? No, you hold on till you find a buyer that appreciates the value of your asset.
There will be a flood of buyers when we have traction or market conditions change and the value of the company is better understood. Whether that’s tomorrow or next month or 2024. You don’t lose unless you sell.
That said, I’m starting out on the scotch tonight. Stay positive and as Gen Z say HODL!
There will be a flood of buyers when we have traction or market conditions change and the value of the company is better understood. Whether that’s tomorrow or next month or 2024. You don’t lose unless you sell.
That said, I’m starting out on the scotch tonight. Stay positive and as Gen Z say HODL!
Wickedwolf
Regular
Jim Cramer almost makes the motley fools look professional
Fullmoonfever
Top 20
Was having a look at the recent TinyML conference which we know Anil was at for the Sept one.
This one is the EMEA Oct one and noticed that one of our Engineers was doing a Demo / Poster as below fwiw as not sure if posted before.
www.tinyml.org
Demos & Posters
Exploiting activation sparsity for easy model optimization in the Akida processor
Sébastien CROUZET, R&D Engineer, Brainchip
Abstract (English)
Neuromorphic processors offer high processing efficiency, but at the cost of an apparent technical barrier for the new user. BrainChip’s Akida processor mitigates this obstacle by running standard convolutional neural networks, and by offering transparent and performant quantization and lossless conversion methods. Neuromorphic efficiency is retained because the event–based processor can exploit the sparsity (high number of zero–values) inherently present in CNN activation maps.
Nonetheless, selection and optimization of models to target Akida (or indeed, any dedicated hardware) can seem daunting without expert knowledge. Here we show that i) by applying simple off–the–shelf optimization techniques, such as regularization and pruning, we can significantly accelerate inference at minimal reduction to accuracy; ii) applying these techniques provides a powerful lever to manipulate the speed/accuracy trade–off according to task requirement; iii) this in turn reduces the requirement to maintain an extensive library of pretrained models to cover a range of task complexities – instead, a handful of pretrained models allows a continuum of final model sizes.
[Method] We select a representative and well–known benchmark task, visual wakewords (VWW), starting from a tiny–ML relevant solution, MobileNet V1, alpha=0.25 (as per the tinyML Perf benchmark). We use a transfer learning–based pipeline, with imagenet pretraining, and quantization of both weights and activations to 4–bits (first layer 8–bit weights), 5 epochs of fine tuning on VWW. We test 2 techniques to optimize performance on our hardware: i) activity regularization added during training (mix of L1 and L2 loss), testing model accuracy and sparsity as a function of the amplitude of regularization ii) structured pruning (whole filters), using a well–known metric for filter selection (batch
normalization gamma). We first test the “prunability” of each layer individually by evaluating model accuracy (without re–training) as a function of % filters pruned. We then test several pruning strategies for the whole model (fixed % accuracy loss per layer; targeted pruning of the high–cost layers; …).
[Results] We first introduce a minor network optimization from the original MobileNet based on known properties of Akida: for low filter numbers (<=64), the efficient implementation of standard convolutional layers in Akida means that they run as fast as an equivalently sized separable convolution, while offering a moderately increased representational capacity and thus accuracy. Specifically, the first 3 separable conv layers of MobileNet v1 are swapped for standard conv layers, forming what we will call “AkidaNet”. With quantization at 4–bits, and using the alpha=0.25 network, we obtain a baseline accuracy of 83.3% on the VWW validation set, with single–sample (batch–size = 1) processing rate of 66 FPS (latency 15.15 ms). Applying a combination of regularization and pruning to optimize this model, we can accelerate processing to 133 FPS while maintaining >80% accuracy. We then test optimization from larger starting models (alpha=0.5 or even 1.0) and find that we can reduce models sufficiently to achieve processing rates similar to the baseline, but with higher accuracy (a=0.5, 84.1% acc. at 84 FPS; a=1.0, 84.8% acc. at 69 FPS)
Also see that TDK InvenSense playing with minimal Bit Weight Quantization as well.
Towards Universal 1-bit Weight Quantization of Neural Networks with end-to-end deployment on ultra-low power sensors
Le MINH TRI, Ph.D. student in deep learning, TDK InvenSense
This one is the EMEA Oct one and noticed that one of our Engineers was doing a Demo / Poster as below fwiw as not sure if posted before.
EMEA 2022 | tinyML Foundation
TinyML Summit. The topic is advances in ultra-low power Machine Learning technologies and applications.
Demos & Posters
Exploiting activation sparsity for easy model optimization in the Akida processor
Sébastien CROUZET, R&D Engineer, Brainchip
Abstract (English)
Neuromorphic processors offer high processing efficiency, but at the cost of an apparent technical barrier for the new user. BrainChip’s Akida processor mitigates this obstacle by running standard convolutional neural networks, and by offering transparent and performant quantization and lossless conversion methods. Neuromorphic efficiency is retained because the event–based processor can exploit the sparsity (high number of zero–values) inherently present in CNN activation maps.
Nonetheless, selection and optimization of models to target Akida (or indeed, any dedicated hardware) can seem daunting without expert knowledge. Here we show that i) by applying simple off–the–shelf optimization techniques, such as regularization and pruning, we can significantly accelerate inference at minimal reduction to accuracy; ii) applying these techniques provides a powerful lever to manipulate the speed/accuracy trade–off according to task requirement; iii) this in turn reduces the requirement to maintain an extensive library of pretrained models to cover a range of task complexities – instead, a handful of pretrained models allows a continuum of final model sizes.
[Method] We select a representative and well–known benchmark task, visual wakewords (VWW), starting from a tiny–ML relevant solution, MobileNet V1, alpha=0.25 (as per the tinyML Perf benchmark). We use a transfer learning–based pipeline, with imagenet pretraining, and quantization of both weights and activations to 4–bits (first layer 8–bit weights), 5 epochs of fine tuning on VWW. We test 2 techniques to optimize performance on our hardware: i) activity regularization added during training (mix of L1 and L2 loss), testing model accuracy and sparsity as a function of the amplitude of regularization ii) structured pruning (whole filters), using a well–known metric for filter selection (batch
normalization gamma). We first test the “prunability” of each layer individually by evaluating model accuracy (without re–training) as a function of % filters pruned. We then test several pruning strategies for the whole model (fixed % accuracy loss per layer; targeted pruning of the high–cost layers; …).
[Results] We first introduce a minor network optimization from the original MobileNet based on known properties of Akida: for low filter numbers (<=64), the efficient implementation of standard convolutional layers in Akida means that they run as fast as an equivalently sized separable convolution, while offering a moderately increased representational capacity and thus accuracy. Specifically, the first 3 separable conv layers of MobileNet v1 are swapped for standard conv layers, forming what we will call “AkidaNet”. With quantization at 4–bits, and using the alpha=0.25 network, we obtain a baseline accuracy of 83.3% on the VWW validation set, with single–sample (batch–size = 1) processing rate of 66 FPS (latency 15.15 ms). Applying a combination of regularization and pruning to optimize this model, we can accelerate processing to 133 FPS while maintaining >80% accuracy. We then test optimization from larger starting models (alpha=0.5 or even 1.0) and find that we can reduce models sufficiently to achieve processing rates similar to the baseline, but with higher accuracy (a=0.5, 84.1% acc. at 84 FPS; a=1.0, 84.8% acc. at 69 FPS)
Also see that TDK InvenSense playing with minimal Bit Weight Quantization as well.
Towards Universal 1-bit Weight Quantization of Neural Networks with end-to-end deployment on ultra-low power sensors
Le MINH TRI, Ph.D. student in deep learning, TDK InvenSense
Kachoo
Regular
Sometimes there are project that have lump some payments and the support is free. You have to remember that this tech is really new to many. That said I will say when BRN give employees shares for milestones what merit are they based on I wish revenue was a measure to pay out employees too. But I don't know there contracts either. Disappointed yes.
Fullmoonfever
Top 20
EOD slightly wider view to engage the bottom possible supp zones.
Similar threads
