BRN Discussion Ongoing

IloveLamp

Top 20
I had already noticed in last month’s job offer for a new HR manager (replacing Sheila Sabanal-Lau) that our company recently implemented cost saving measures by limiting the legendary free lunch to 2-3 times a week! 😂

Apparantly now on those 3 days its all you can eat!
Good. I'm glad they're treating the staff well. A free lunch is peanuts compared to what some companies are offering.

When we start making consistent $$, i hope to see the wages increase in order to remain competitive and not bleed staff to the nvidias and intels of the world
 
Last edited:
  • Like
  • Fire
Reactions: 9 users

Kachoo

Regular
Good. I'm glad they're treating the staff well. A free lunch is peanuts compared to what some companies are offering.

When we start making consistent $$, i hope to see the wages increase, in order to remain competitive and not bleed staff to the nvidias and intels of the world
I don't know if many realise but if you provide lunches say lunch is 1 hour usually most people will be ba k to work 45 minutes or even 40 mins.

Where if you go out its usually 1.5 hours so when you look at the rates you pay a worker even 15 minutes covers the cost of the lunch to increase productivity.

Is a very smart move.
 
  • Like
  • Haha
Reactions: 8 users

Frangipani

Top 20
I had already noticed in last month’s job offer for a new HR manager (replacing Sheila Sabanal-Lau) that our company recently implemented cost saving measures by limiting the legendary free lunch to 2-3 times a week! 😂

Apparantly now on those 3 days its all you can eat!

… and doggy-bag the leftovers for one of the other days? 🤣

Actually, for the new HR manager, it’s free lunches on Mondays and Fridays only:

16465C89-F564-430A-A3E0-B3ADEABB2958.jpeg



While it’s still a nice gesture to provide free lunch twice a week, it seems a bit odd to cut down on the free lunches for new employees (unless they also did so for everyone else), especially when paying them a salary of more than 110k USD a year!
 
  • Like
  • Love
Reactions: 3 users
The free lunch on Monday and Friday would be to entice them to come in to the office, instead of having a long weekend and say they plan to work from home on those days.
 
  • Like
  • Fire
Reactions: 13 users

BrainShit

Regular
Why would Loihi scale better than Akida? Akida being available as IP puts it in a scaling league of its own don’t you think? The node offerings mean it can be embedded as small as two nodes or as large as 256 nodes. I don’t think BrainChip have ever left Akida in the Data Centre off the table either. Versatility is definitely a thing being able to offer up both hardware and software versions of both TENNs and Akida - something that I’m pretty sure Loihi is not capable of. TENNs makes our product line extremely agile imo.


Yes, you're right...

Just regarding scaling (not size, weight, power-consumption or body fat and available + proofed)... I compared 256 nodes which will scale up to 131 TOPS with Hala Point (Loihi2 cluster) which lead to 20 petaops ...

I know this comparison is not well suited, but it shows the dimension as well as the scaling ability of Loihi 2. So, next step ... who like to set up an Akida cluster?

Link: https://www.nextplatform.com/2024/0...ic-ai-envelope-with-hala-point-supercomputer/

Link: https://brainchip.com/wp-content/uploads/2022/06/Akida-1.0-IP-Product-Brief_final.pdf


We're on the same ground... Akida is way better than Loihi and Intel grind's its teeth...
 
Last edited:
  • Like
  • Fire
Reactions: 10 users
Seems Microsoft maybe seeing the light :unsure:

I note earlier this year they apparently signed a deal with Intel to manufacture their chips, announced during the Intel Foundry event. Presumably the same one we attended & showcased at.

New recent Intern opening with an excerpt, my bold.

Intels neuromorphic or how good would it be for them to maybe use an IFS partner :)


Research Intern - AI-driven Hardware Design​

Microsoft · 6 days ago · Vancouver, British Columbia, Canada

Preferred Qualifications • Proficient analytical and problem-solving skills and communication skills, both written and verbal. • Ability to work independently and collaboratively in a dynamic research environment. • Proficient understanding of AI and machine learning concepts, especially in relation to hardware infrastructure. • Experience in simulation modeling and software development (e.g., Python, C++, MATLAB). • Research expertise with AI hardware components such as GPUs, TPUs, and neuromorphic chips. Hardware RTL development (SystemVerilog, Chisel, Bluespec) is a plus. • Experience with machine learning frameworks (TensorFlow, PyTorch). • Deep Knowledge of hardware design and architecture. • Prior research or project experience in AI or hardware simulation.
 
  • Like
  • Love
  • Fire
Reactions: 23 users

Sirod69

bavarian girl ;-)
On LinkedIn:
BrainChip's Akida technology is inspired by the remarkable capabilities of the human brain in processing and learning. Positioned at the forefront of innovation, our technology is set to dynamically transform the way systems and applications process information, unlocking unprecedented efficiency and intelligence. Learn more: www.brainchip.com
1717087872793.png
 
  • Like
  • Love
  • Fire
Reactions: 32 users
The free lunch on Monday and Friday would be to entice them to come in to the office, instead of having a long weekend and say they plan to work from home on those days.
Did someone say free lunch

1717092611994.gif
 
  • Haha
  • Wow
  • Like
Reactions: 6 users

yogi

Regular
sorry if posted already

 
  • Like
  • Love
  • Fire
Reactions: 37 users

Damo4

Regular
 
Last edited:
  • Like
  • Thinking
  • Fire
Reactions: 33 users

Esq.111

Fascinatingly Intuitive.
Morning Damo 4 ,

Must have missed this one.... 20 / 5 / 24.???

As a engineer years ago , one of the first things I was taught ... measure twice , cut once.

Esq.
 
  • Like
  • Fire
Reactions: 12 users

Damo4

Regular
Morning Damo 4 ,

Must have missed this one.... 20 / 5 / 24.???

As a engineer years ago , one of the first things I was taught ... measure twice , cut once.

Esq.
Updated now with the corrected tweet
 
  • Like
  • Haha
  • Fire
Reactions: 7 users

Evermont

Stealth Mode
  • Like
  • Fire
  • Love
Reactions: 19 users
  • Like
  • Love
Reactions: 30 users

Dougie54

Regular
I had already noticed in last month’s job offer for a new HR manager (replacing Sheila Sabanal-Lau) that our company recently implemented cost saving measures by limiting the legendary free lunch to 2-3 times a week! 😂

Apparantly now on those 3 days its all you can eat!
Another nose for the feed bag
 
  • Like
  • Fire
  • Thinking
Reactions: 7 users
 
  • Like
  • Fire
  • Love
Reactions: 11 users
When putting Neuromorphic into the search bar of Accenture website there a four videos on the topic, here is one


 
  • Like
  • Thinking
  • Fire
Reactions: 10 users
Just a guess but maybe we might see a few lines swiped this afternoon.

1717123295079.gif
 
  • Like
  • Haha
  • Fire
Reactions: 6 users

jtardif999

Regular
Good evening...here we go again, yet another patent only published 7 days ago from our mates at:

ACCENTURE GLOBAL SOLUTIONS LTD

Check it out, a research chip is mentioned but a commercially available chip or IP is available NOW...that's why the
continuous learning of gesture recognition and AKIDA IS MENTIONED YET AGAIN.



Patents are very powerful tools, researching them is a must !

Tech ;)
Thanks again @TECH - from the description and claims this can only be Akida. Accenture must have some pretty big customer(s) for all these patents.

I particularly like the following description: “if a user corrects the gesture recognition model by, for example, vocally giving a description of the gesture that the model misrecognized through a microphone of the device, a conversational agent of the human machine interface can trigger a learning process of the neuromorphic processor and the neuromorphic processor can update the gesture recognition model based on the correction in real-time, e.g., within seconds, of the correction being detected. This real-time updating reduces the number of mistakes made by a gesture recognition component of the device in the future and results in faster and more accurate learning of the gesture recognition model, which in turn reduces user frustration and reduces the number of errors caused by misrecognizing user gestures that cause other components to perform actions.”
 
  • Like
  • Fire
  • Love
Reactions: 36 users
Top Bottom