BRN Discussion Ongoing

Diogenese

Top 20
Great article FMF but I am sort of happily confused by the following paragraph:

“New DesignWare ARC NPX6 NPU IP​

To keep pace with the evolving neural network advancements and the growing demand for higher performance, Synopsys has recently introduced the DesignWare® ARC® NPX6 (Fig 4) NPU IP. The NPX6 NPU IP addresses demands of real-time compute with ultra-low power consumption for deep learning applications. The NPX6 NPU IP is Synopsys’ sixth generation neural network accelerator IP.”

Synopsys claims it’s own neural network accelerator IP yet chooses to use and thereby advertise and by implication endorse/recommend Brainchip’s artificial intelligence accelerator by use of the photo in its report.

A curious turn of events worthy of a novel by Lewis Carroll.

My opinion only DYOR
FF

AKIDA BALLISTA

Synopsys are mainly about the provision of circuit design IP tools, rather than the actual SoC.

Their Products page talks of IP solutions:

https://www.synopsys.com/dw/ipdir.php?ds=arc-npx6#

DesignWare® ARC® NPX Neural Processor IP family provides a high-performance, power- and area-efficient IP solution for a range of applications requiring AI enabled SoCs. The ARC NPX6 NPU IP is designed for deep learning algorithm coverage including both computer vision tasks such as object detection, image quality improvement, and scene segmentation, and for broader AI applications such as audio and natural language processing.

https://www.synopsys.com/silicon-design.html

The Silicon Powering the Software

Advanced silicon chips power the amazing software we rely on every day. They are the foundation for everything from smartphones and wearables to self-driving cars and machines that learn. Synopsys is the leader in solutions for designing and verifying complex chips and for designing the advanced processes and models required to manufacture those chips.


They are seemingly still dealing with CNNs, and do not appear to be involved with spiking NNs.

https://www.synopsys.com/dw/ipdir.php?ds=arc-npx6

DesignWare ARC NPX6 NPU Family for AI / Neural Processing​



DesignWare® ARC® NPX Neural Processor IP family provides a high-performance, power- and area-efficient IP solution for a range of applications requiring AI enabled SoCs. The ARC NPX6 NPU IP is designed for deep learning algorithm coverage including both computer vision tasks such as object detection, image quality improvement, and scene segmentation, and for broader AI applications such as audio and natural language processing.
The NPX6 NPU family offers multiple products to choose from to meet your specific application requirements. The architecture is based on individual cores that can scale from 4K MACs to 96K MACs for a single AI engine performance of over 250 TOPS and over 440 TOPS with sparsity. The NPX6 NPU IP includes hardware and software support for multi-NPU clusters of up to 8 NPUs achieving 3500 TOPS with sparsity. Advanced bandwidth features in hardware and software, and a memory hierarchy (including L1 memory in each core and a high-performance, low-latency interconnect to access a shared L2 memory) make scaling to a high MAC count possible. An optional tensor floating point unit is available for applications benefiting from BF16 or FP16 inside the neural network.
To speed application software development, the ARC NPX6 NPU Processor IP is supported by the MetaWare MX Development Toolkit, a comprehensive software programming environment that includes a neural network Software Development Kit (NN SDK) and support for virtual models. The NN SDK automatically converts neural networks trained using popular frameworks, like Pytorch, Tensorflow, or ONNX into optimized executable code for the NPX hardware.
The NPX6 NPU Processor IP can be used to create a range of products – from a few TOPS to 1000s of TOPS – that can be programmed with a single toolchain.

Highlights
  • Scalable real-time AI / neural processor IP with up to 3,500 TOPS performance
  • Supports CNNs, RNNs/LSTMs, transformers, recommender networks, etc.
  • Industry leading power efficiency (up to 30 TOPS/W)
  • 1-24 cores of an enhanced 4K MAC/core convolution accelerator
  • Tensor accelerator providing flexible activation and support of Tensor Operator Set Architecture (TOSA)
  • Software Development Kit
    • Automatic mixed mode quantization tools
  • Bandwidth reduction through architecture and software tool features
  • Latency reduction through parallel processing of individual layers
  • Seamless integration with DesignWare ARC VPX vector DSPs
  • High productivity MetaWare MX Development Toolkit supports Tensorflow and Pytorch frameworks and ONNX exchange format


1662548292177.png



US10846591B2 Configurable and programmable multi-core architecture with a specialized instruction set for embedded application based on neural networks

A programmable architecture specialized for convolutional neural networks (CNNs) processing such that different applications of CNNs may be supported by the presently disclosed method and apparatus by reprogramming the processing elements therein. The architecture may include an optimized architecture that provides a low-area or footprint and low-power solution desired for embedded applications while still providing the computational capabilities required for CNN applications that may be computationally intensive, requiring a huge number of convolution operations per second to process inputs such as video streams in real time.

1662550568062.png


Interestingly, the claims refer to "A CNN architecture ..." rather than "A CNN processor ..." or similar.


However, given Synopsys' use of the BrainChip Accelerator photo, one possible scenario is that BrainChip used some earlier version of their design tools to plot the layout of the Accelerator onto a FPGA all those years ago, because not even Anil does IC layouts by hand (late edition - "except maybe the fine tuning").

https://brainchip.com/brainchip-introduces-worlds-first-commercial-hardware-acceleration-of-neuromorphic-computing-brainchip-120917/#:~:text=As the first commercial implementation of a hardware-accelerated,branch of artificial intelligence that simulates neuron functions.

As the first commercial implementation of a hardware-accelerated spiking neural network system, BrainChip Accelerator is a significant milestone in the development of neuromorphic computing, a branch of artificial intelligence that simulates neuron functions. The processing is done by six BrainChip Accelerator cores in a Xilinx Kintex Ultrascale field-programmable gate array (FPGA).
 
Last edited:
  • Like
  • Love
  • Fire
Reactions: 26 users

uiux

Regular
nppus68.png
 
  • Like
  • Haha
  • Fire
Reactions: 23 users

Diogenese

Top 20
  • Like
  • Love
  • Haha
Reactions: 9 users

Bravo

If ARM was an arm, BRN would be its biceps💪!
A tiny update on the Brainchip Synopsys link.

I have had confirmed unofficially that Synopsys is a supplier to Brainchip. I do not know what they supply. I may receive more unofficial information but unofficially it was agreed that it could be seen as strange that as a supplier only that they would use this image in their report.

My official position is that I have no doubt that it is definitely not by happenstance.

My opinion only DYOR
FF

AKIDA BALLISTA

Hi FF,

I've been wondering whether Synopsys could have something to do with the compliance element, because they provide IP solutions for SoC designs, and they also make them comply with ISO 26262.

Just a thought.
 
  • Like
  • Thinking
Reactions: 9 users
Hi FF,

I've been wondering whether Synopsys could have something to do with the compliance element, because they provide IP solutions for SoC designs, and they also make them comply with ISO 26262.

Just a thought.
Hi @Bravo
I think the odds are against this idea as the engagement between Brainchip and Synopsys predates Anil Mankar’s 2021 Ai Field Day presentation where he was asked about ISO verification and he was adamant Brainchip was not interested in undertaking it because of cost and time and they were leaving it to the OEM’s and automotive customers they were engaged with to undertake.

My opinion only DYOR
FF

AKIDA BALLISTA
 
  • Like
  • Fire
  • Love
Reactions: 9 users

Proga

Regular
This is pretty huge. Can you see the convergence. everything is coming together. MegaChips is going to be able to offer complete solutions for billions of low powered internet enabled edgeAI devices in all locations (including remote places) enabling Akida's low powered sensor fusion to operate securely and when necessary return actionable information to centralised GPU. Our future edgeAI wearables will be able to stay connected, call for help, provide updated advice from other devices.

Remember when Qualcomm preached the future of everything connected to everything. This technology will be part of that future.

Autonomous driving is smarter and safer and faster from the live updates on the ground shared with edge AI devices powered by Akida.
I can envisage a hefty chunk of our revenue generated from MegaChip customers
 
  • Like
  • Love
  • Fire
Reactions: 9 users
D

Deleted member 118

Guest
Just arrived home to Australia and found this an interesting article to read. Will it or will it not affect BRN in the future? As I can’t see the American Government being to pleased with any technology getting into chinas hands let alone Akida


US chip makers hit by new China export rule
By Annabelle Liang
Business reporter
Published
6 days ago

Share
The world's first AI general infrastructure system built on NVIDIA A100 chips on display in Hangzhou China.
IMAGE SOURCE, GETTY IMAGES
Image caption,
Nvidia is a top manufacturer of computer chips
Shares of major chipmakers Nvidia and AMD have fallen amid concerns of new US restrictions on the sale of artificial intelligence chips to China.
Nvidia says the US government requires a new licence, effective immediately, to address the risk of chips being "used in, or diverted to a 'military end use'... in China and Russia".
There are fears the rule could lead to millions of dollars in lost revenue.
Shares of both chipmakers slipped in after-hours trading in New York.
Nvidia's shares were down by 6.6% while AMD slipped 3.7%.
ADVERTISEMENT

The new restrictions are a "gut punch for Nvidia", Dan Ives of Wedbush Securities told the BBC.
US tries to take on China with huge tech investment
Why is there a chip shortage?
Chinese officials have firmly opposed the latest move. According to state media, "actions from the United States deviated from the principle of fair competition and violated international economic and trade rules".
In a statement, Beijing said "The US side should immediately stop its wrongdoing, treat companies from all over the world including Chinese companies fairly, and do more things that are conducive to the stability of the world economy."
ADVERTISEMENT


The US Commerce Department told the BBC it was "not in a position to outline specific policy changes at this time".
"We are taking a comprehensive approach to implement additional actions necessary related to technologies, end-uses, and end-users to protect US national security and foreign policy interests," a Commerce Department spokesperson said.
"This includes preventing China's acquisition and use of US technology in the context of its military-civil fusion program to fuel its military modernisation efforts, conduct human rights abuses, and enable other malign activities."
In a US regulatory filing on Wednesday, Nvidia said the new licence requirement would hit exports of its A100 and H100 chips, which are designed to speed up machine learning tasks, and the systems which include them.
Around $400m (£345.2m) in sales to China could be affected, Nvidia added, "if customers do not want to purchase the company's alternative product offerings or if the (US government) does not grant licenses in a timely manner or denies licenses to significant customers".
A Nvidia spokesperson told BBC it was liaising with customers in China "to satisfy their planned or future purchases with alternative products".
Meanwhile, an AMD spokesperson said the rules, which would prevent the shipment of its MI250 chips to China, were not expected to have "a material impact" on business.
Both Nvidia and AMD halted sales to Russia after the invasion of Ukraine in February.
Analysts said the US requirements could make it more difficult for China to acquire chips for advanced computing.
It could also affect the earnings of US manufacturers such as Nvidia and AMD, said Mario Morales, a California-based analyst at market intelligence firm IDC.
"Both companies have a large exposure to China and could see more impact going forward, especially if China chooses to retaliate," Mr Morales said.
Rising tensions
Last week, Nvidia reported a revenue of $6.7bn in the second quarter, which was significantly lower than forecasts.
However, it said revenue from its data centre business - which produces computer chips - surged by 61% from a year earlier.
"This is really a shot across the bow at China and it's really going to fan those flames in terms of geopolitical (tensions). Nvidia's caught in the crossfire," Mr Ives said.
The US and China are locked in a long-running dispute over trade and technology.
Tensions between the world's two biggest economies rose earlier this month, after US politician Nancy Pelosi made a controversial visit to Taiwan.
China sees the self-ruled island as a part of its territory and insists it should be unified with the mainland, by force if necessary.
 
Last edited by a moderator:
  • Like
Reactions: 7 users

Proga

Regular
South Korea’s Samsung Electronics Co. is warning that the semiconductor industry could be in for a rocky close to 2022.

A senior executive at the world’s largest maker of memory chips said the outlook for the second half of the year is gloomy, and Samsung is not yet seeing momentum for a recovery next year. Rival chipmakers such as SK Hynix Inc. and Micron Technology Inc. have cautioned about slowing demand in recent weeks.

“The general perception earlier this year was that the second half of would be better than the first half, but from April to May, it changed drastically,” said Kyung Kyehyun, head of Samsung’s Device Solutions Division, which oversees the company’s semiconductor operations. “The world is changing so quickly.”

Kyung made the comments during a rare briefing at the company’s new chip fab in Pyeongtaek on Wednesday. Samsung’s strategy is to respond faster to market changes, rather than stick to an investment plan prepared in advance, Kyung said during the event. That said, the company will do its best to keep capital expenditures steady, he added.

Samsung historically has invested heavily in new chip initiatives, which now include the foundry business to better compete with Taiwan Semiconductor Manufacturing Co. for global customers. Samsung kicked off mass production of 3-nanometer chips at its foundry in June, edging out TSMC in a race to build the most advanced chips in the world. Samsung will work on improving the performance and lowering the cost of the chips, as it aims to create its next-generation 3 nm chips in 2024, Kyung said.

Besides a slumping chip market, Samsung is also struggling with the clash between China and the US. While South Korea has historically aligned with Washington, the tech giant counts on being able to sell chips, smartphones and other products into the massive Chinese market. Samsung has both customers and factories in China.

“It is difficult for us to miss such a market, and there are many important customers,” said Kyung. “We’re trying to find a win-win solution for everyone in the midst of this conflict.”



The US government is tightening flows of technologies to China, most recently restricting sales of artificial intelligence chips and cutting-edge chip gear to Chinese customers. It is also considering moves to restrict US investment in Chinese tech companies, while at the same time offering billions of dollars in incentives to bolster semiconductor production on American soil. Washington is demanding that any chipmaker receiving a part of the federal grant refrain from manufacturing advanced chips in China for a period of ten years. The Korean government is seeking to negotiate that with US officials.

As the US beefs up efforts to solidify a chip supply chain at home, Samsung announced plans for an advanced $17 billion chip plant in Taylor, Texas, with construction slated to start later this year.

The new plant would lure new clients with closer partnerships, Kyung said, adding the company will continue to invest in the US. Samsung has also floated the idea of a broad expansion of its semiconductor manufacturing facilities in Texas, laying out potential plans to spend almost $200 billion on 11 plants in a series of filings in the state in July.

Seoul is joining working-level talks with US, Taiwan and Japan to explore ways to further corral China’s ambition become a world’s leader in chip technology and lower its dependence on the West.

On Washington’s initiative, dubbed the Chip 4 alliance, Kyung said he hoped South Korea will “seek understanding from China first, and then negotiate with the US.” But he also said, “In the long run, it may be difficult to put new equipment into our fabs in China.”

 
  • Like
  • Fire
Reactions: 8 users
D

Deleted member 118

Guest
I never realised nviso are working with Tesla


338D4168-59E9-418C-9520-F07BC5931C16.png




Experience

CEO and Co-founder
NVISO Human Behaviour AI
Oct 2009 - Present 13 yrs
Lausanne Area, Switzerland
NVISO is humanising autonomous machines for automotive, medical and consumer electronic device manufacturers. We are a leading human behaviour AI software companies providing AI Apps and Solutions with superior "on-device" recognition accuracy with massive real-world database (data is one key pillar for AI). Key applications in making autonomous vehicles (e.g. Tesla) safe with interior monitoring, patient monitoring for remote tele-medicine in healthcare/ robotics, and intelligent devices for smart homes. Proprietary technology developed over 10 years with collaborative R&D funding in collaboration with Europe’s leading AI researchers (EPFL, ETHZ, University of Edinburgh, Technical University of Munich). We focus on deployment to edge AI devices such as ARM, Intel, NVIDIA and more deeply embedded systems of TI, Renesas, Qualcomm, etc. Our clients are include Fortune 500 corporations to ground breaking start-ups. We have offices in India, Serbia, Japan, and Australia.
 

Attachments

  • 81B6217E-FCBA-4986-BA75-E1BDA6626338.png
    81B6217E-FCBA-4986-BA75-E1BDA6626338.png
    1.6 MB · Views: 62
  • Like
  • Fire
  • Thinking
Reactions: 27 users

MDhere

Regular
I don't assume Akida has also been chosen... but ..I intuit it. I have a very strong natural instinct for intuition, similar to a pig finding truffles.
what month and year is yr intuition that sp is $10?
 
  • Haha
  • Like
Reactions: 6 users
It also most likely ... "unofficially official that" ...the next 4C revenue figures will be significant and speak for themselves ...!!
Hopefully the share price might reflect a better quarterly report
 
  • Like
Reactions: 4 users

clip

Regular
  • Like
  • Fire
  • Love
Reactions: 11 users
 
  • Like
  • Fire
  • Love
Reactions: 17 users

Potato

Regular
Screen Shot 2022-09-08 at 4.57.42 am.png
 
  • Like
  • Fire
  • Love
Reactions: 28 users

Potato

Regular
  • Like
  • Fire
  • Love
Reactions: 10 users

Lex555

Regular
I never realised nviso are working with Tesla


View attachment 16088



Experience

CEO and Co-founder
NVISO Human Behaviour AI
Oct 2009 - Present 13 yrs
Lausanne Area, Switzerland
NVISO is humanising autonomous machines for automotive, medical and consumer electronic device manufacturers. We are a leading human behaviour AI software companies providing AI Apps and Solutions with superior "on-device" recognition accuracy with massive real-world database (data is one key pillar for AI). Key applications in making autonomous vehicles (e.g. Tesla) safe with interior monitoring, patient monitoring for remote tele-medicine in healthcare/ robotics, and intelligent devices for smart homes. Proprietary technology developed over 10 years with collaborative R&D funding in collaboration with Europe’s leading AI researchers (EPFL, ETHZ, University of Edinburgh, Technical University of Munich). We focus on deployment to edge AI devices such as ARM, Intel, NVIDIA and more deeply embedded systems of TI, Renesas, Qualcomm, etc. Our clients are include Fortune 500 corporations to ground breaking start-ups. We have offices in India, Serbia, Japan, and Australia.
Do you have any other documentation, only thing mentioned of Tesla here, is him is using it as an example of a company requiring autonomous applications
 
  • Like
  • Love
Reactions: 11 users
D

Deleted member 118

Guest
Do you have any other documentation, only thing mentioned of Tesla here, is him is using it as an example of a company requiring autonomous applications
Weird why he mentioned Tesla.
 
Last edited by a moderator:
  • Like
Reactions: 2 users
Interesting:

From their website:

“Artificial IntelligentWe find the path for a safer life with AI.
From detection to tracking and classification, the AI-based core algorithm of Vueron enables the 3D point cloud to transform accurate and highly valuable information into useful data for a comfortable life”

 
  • Like
  • Fire
Reactions: 11 users
Top Bottom