BRN Discussion Ongoing
- Thread starter TechGirl
- Start date
Hi jtardif999,I’m going to mess up your math FF. Akida1.0 has 80 nodes and each node has 4 NPUs. Everywhere you refer to the number of NPUs can be multiplied by 4
The Akida 1000 SoC has 20 nodes, each node has 4 NPUs, but this is somewhat academic, as we no longer make the SoC. The Akida IP can be implemented in 2 nodes, up to 256 nodes.
Products - BrainChip
https://brainchip.com/products/
Download the Akida 1.0 product brief:
Scale down to 2 nodes (@ 1Ghz = 1 TOPS) for ultra low power or scale up to 256 nodes (@ 1Ghz = 131 TOPS) for complex use cases.
...
Every node consists of four Neural Processing Units (NPUs), each with scalable and configurable SRAM. Within each node, the NPUs can be configured as either convolutional or fully connected. The Akida neural processor is event based – leveraging data sparsity, activations, and weights to reduce the number of operations by at least 2X.
Note that the web page from which the product brief can be downloaded contains an inconsistent use of the word "node" in that it suggests that 1024 nodes are possible, whereas the Product brief makes it clear that 256 nodes is the maximum:
"Infer and learn at the edge with Akida’s fully customizable event-based AI neural processor. Akida’s scalable architecture and small footprint boosts efficiency by orders of magnitude – supporting up to 1024 nodes that connect over a mesh network."
Xhosa12345
Regular
Nup, more sci fi, this is where the big bucks are....!!
Although money is not really a thing in the federation... but ill be happy with some bars of gold pressed latinum !
Boab
I wish I could paint like Vincent
Am I reading it wrong. Each Unit costs around $7-10 with 2500 units around $20,000??
VictorG
Member
No, i read it wrong. I suck at being a shorter.
Fact Finder
Top 20
No you are reading it correctly across all the distributors average cost is around $US8.00.
Nviso was running multiple Apps doing lots of in cabin monitoring of all passengers and using only 5% of AKIDA 1.0 capacity.
On this basis just driver fatigue and distraction would be easily run on a couple of nodes which is 8 NPEs so a car company could pick up AKIDA IP and Nviso for less than $1.50 possibly $1.00.
Eight times less costly but with the additional opportunity to upgrade in the various makers models right up through Nviso’s 20 or 24 Apps which include medical health monitoring as well as passenger monitoring etc;
Brainchip AKIDA is killer Science Fiction.
My opinion only DYOR
FF
AKIDA BALLISTA
Boab
I wish I could paint like Vincent
Many thanks for the explanation.
I have to remember Akida can do many things at the same time.
Cheers
I did some research and Renasas is supplying chips to Syntiant
Getupthere
Regular
Perceive AI Launches 2nd Edge AI Chip For Low Power Applications — Forbes
Company claims Ergo2 is up to four times faster than Perceive’s first-generation Ergo chip, and can handle much larger models such as NLP.
Perceive AI Launches 2nd Edge AI Chip For Low Power Applications
Company claims Ergo2 is up to four times faster than Perceive’s first-generation Ergo chip, and can handle much larger models such as NLP.
Edge AI is coming into its own, with a variety of chips being launched that offer low cost, low power, and high performance. While training AI models gets most of the attention in the media, Inference processing will end up getting most of the revenue, especially at the edge. Global Market Insights, a respected market intelligence firm, projects that the market for Edge AI will top $5B in 2023, with a 20% CAGR through the next decade. While our gut feel is that $5B is too high, we feel the 20% growth forecast is far too low.
At any rate, the market is attracting many contenders, including Perceive, which was spun out of Xperi Corporation in 2018 to concentrate on this opportunity, and now already has its second product ready for market.
What did Perceive Announce?
The company is not replacing its Ergo product, but rather is adding a higher performance and more capable chip for demanding edge applications. As the table below shows, the new device delivers a big boost in image classification, and consumes less than 20 mW. Thats Milliwatts, or 1/1000 of a watt. We know of no competitor who can claim that, and still deliver about 1000 inferences per second.
There are many companies readying or shipping chips for Edge AI, including SiMa.ai, Hailo Technologies, AlphaICs, Recogni, EdgeCortix, Flex Logix, Roviero, BrainChip, Syntiant, Untether AI, Expedera, Deep AI, Andes, Plumerai, in addition to Intel, AMD (Xilinx) and of course NVIDIA. Some, like NVIDIA and SiMa.ai are heading down the SoC route, where the chip offers a more complete solution including Arm or RISC-V CPU cores and I/O.
In contrast, Perceive (and others such as Hailo) has focussed on customers who are looking for an AI accelerator that attaches to an SoC for a specific application. Interestingly, the Ergo chip does not require external DRAM, although it supports connectivity to a NOR Flash to contain the weights for larger models. This can be a cost advantage for applications such as speech-to-text , audio applications, and video processing tasks like video super resolution and pose detection. Compared to existing products such as the Hailo-8 accelerator at 2-4 watts, the Ergo is targeting lower power (tens of milliwatts vs 2-4 watts for Hailo-8) albeit with lower performance.
Conclusions
As we’ve always said, the edge market is far easier to penetrate than training in the data center because different applications have dramatically different requirements. While an image processor for autonomous vehicles requires higher performance at higher power levels, a smart camera or an embedded speech to text processor demands a lower power envelope, and a lower cost. And many auto makers will prefer an SoC like NVIDIA Drive instead of designing their own SoC; Tesla is an exception to the rule.
Consequently, there is plenty of room for specific low-power processors such as Perceive, and the company is smart to quickly extend their first foray with a faster sub-watt processor. The software will be key to their success, enabling larger models to run efficiently on the new Ergo2.
We would note that there are many more competitors entering this market, so check back soon to stay up to date!
C-Suite News, Analysis And Advice For Top Decisionmakers
Sign up for the Forbes CxO Newsletter sent every Monday.
Can't wait for Sean Hehir's next CEO address at the AGMImagine if BRN all of a sudden released its 4c now rather than late January and it contained good revenue. Shorters would get destroyed.
David holland
Regular
I can't either the CEO Made a statement, and I want to see if his backed up his words, no excuses if you put your head on the block pls make sure your rightCan't wait for Sean Hehir's next CEO address at the AGM
Dhm
Regular
It is articles like this one that hide Brainchip behind a veil of perceived mediocracy. A whole heap of 'competitors' in edge ai inference make a reader believe that the edge market is populated with like or similar products. It doesn't, for example, say that Intel pushed its own product aside to embrace Akida. I would love it for someone with the skill set to do a table of competitors to compare just who is the undisputed leader in this edge field.
Perceive AI Launches 2nd Edge AI Chip For Low Power Applications — Forbes
Company claims Ergo2 is up to four times faster than Perceive’s first-generation Ergo chip, and can handle much larger models such as NLP.apple.news
Perceive AI Launches 2nd Edge AI Chip For Low Power Applications
Company claims Ergo2 is up to four times faster than Perceive’s first-generation Ergo chip, and can handle much larger models such as NLP.
Edge AI is coming into its own, with a variety of chips being launched that offer low cost, low power, and high performance. While training AI models gets most of the attention in the media, Inference processing will end up getting most of the revenue, especially at the edge. Global Market Insights, a respected market intelligence firm, projects that the market for Edge AI will top $5B in 2023, with a 20% CAGR through the next decade. While our gut feel is that $5B is too high, we feel the 20% growth forecast is far too low.
At any rate, the market is attracting many contenders, including Perceive, which was spun out of Xperi Corporation in 2018 to concentrate on this opportunity, and now already has its second product ready for market.
What did Perceive Announce?
The company is not replacing its Ergo product, but rather is adding a higher performance and more capable chip for demanding edge applications. As the table below shows, the new device delivers a big boost in image classification, and consumes less than 20 mW. Thats Milliwatts, or 1/1000 of a watt. We know of no competitor who can claim that, and still deliver about 1000 inferences per second.
There are many companies readying or shipping chips for Edge AI, including SiMa.ai, Hailo Technologies, AlphaICs, Recogni, EdgeCortix, Flex Logix, Roviero, BrainChip, Syntiant, Untether AI, Expedera, Deep AI, Andes, Plumerai, in addition to Intel, AMD (Xilinx) and of course NVIDIA. Some, like NVIDIA and SiMa.ai are heading down the SoC route, where the chip offers a more complete solution including Arm or RISC-V CPU cores and I/O.
In contrast, Perceive (and others such as Hailo) has focussed on customers who are looking for an AI accelerator that attaches to an SoC for a specific application. Interestingly, the Ergo chip does not require external DRAM, although it supports connectivity to a NOR Flash to contain the weights for larger models. This can be a cost advantage for applications such as speech-to-text , audio applications, and video processing tasks like video super resolution and pose detection. Compared to existing products such as the Hailo-8 accelerator at 2-4 watts, the Ergo is targeting lower power (tens of milliwatts vs 2-4 watts for Hailo-8) albeit with lower performance.
Conclusions
As we’ve always said, the edge market is far easier to penetrate than training in the data center because different applications have dramatically different requirements. While an image processor for autonomous vehicles requires higher performance at higher power levels, a smart camera or an embedded speech to text processor demands a lower power envelope, and a lower cost. And many auto makers will prefer an SoC like NVIDIA Drive instead of designing their own SoC; Tesla is an exception to the rule.
Consequently, there is plenty of room for specific low-power processors such as Perceive, and the company is smart to quickly extend their first foray with a faster sub-watt processor. The software will be key to their success, enabling larger models to run efficiently on the new Ergo2.
We would note that there are many more competitors entering this market, so check back soon to stay up to date!
C-Suite News, Analysis And Advice For Top Decisionmakers
Sign up for the Forbes CxO Newsletter sent every Monday.
I wish I did have such a skillset.
This list of other edge inhabitants contains names I have never heard of. Can those with greater knowledge name which of these names that are pretenders, also those that are coming close, and those that are legitimate competition?
SiMa.ai, Hailo Technologies, AlphaICs, Recogni, EdgeCortix, Flex Logix, Roviero, Syntiant, Untether AI, Expedera, Deep AI, Andes, Plumerai, in addition to Intel, AMD (Xilinx)
“and consumes less than 20 mW. Thats Milliwatts, or 1/1000 of a watt. We know of no competitor who can claim that, and still deliver about 1000 inferences per second.”Perceive AI Launches 2nd Edge AI Chip For Low Power Applications — Forbes
Company claims Ergo2 is up to four times faster than Perceive’s first-generation Ergo chip, and can handle much larger models such as NLP.
Perceive AI Launches 2nd Edge AI Chip For Low Power Applications
Company claims Ergo2 is up to four times faster than Perceive’s first-generation Ergo chip, and can handle much larger models such as NLP.
Edge AI is coming into its own, with a variety of chips being launched that offer low cost, low power, and high performance. While training AI models gets most of the attention in the media, Inference processing will end up getting most of the revenue, especially at the edge. Global Market Insights, a respected market intelligence firm, projects that the market for Edge AI will top $5B in 2023, with a 20% CAGR through the next decade. While our gut feel is that $5B is too high, we feel the 20% growth forecast is far too low.
At any rate, the market is attracting many contenders, including Perceive, which was spun out of Xperi Corporation in 2018 to concentrate on this opportunity, and now already has its second product ready for market.
What did Perceive Announce?
The company is not replacing its Ergo product, but rather is adding a higher performance and more capable chip for demanding edge applications. As the table below shows, the new device delivers a big boost in image classification, and consumes less than 20 mW. Thats Milliwatts, or 1/1000 of a watt. We know of no competitor who can claim that, and still deliver about 1000 inferences per second.
There are many companies readying or shipping chips for Edge AI, including SiMa.ai, Hailo Technologies, AlphaICs, Recogni, EdgeCortix, Flex Logix, Roviero, BrainChip, Syntiant, Untether AI, Expedera, Deep AI, Andes, Plumerai, in addition to Intel, AMD (Xilinx) and of course NVIDIA. Some, like NVIDIA and SiMa.ai are heading down the SoC route, where the chip offers a more complete solution including Arm or RISC-V CPU cores and I/O.
In contrast, Perceive (and others such as Hailo) has focussed on customers who are looking for an AI accelerator that attaches to an SoC for a specific application. Interestingly, the Ergo chip does not require external DRAM, although it supports connectivity to a NOR Flash to contain the weights for larger models. This can be a cost advantage for applications such as speech-to-text , audio applications, and video processing tasks like video super resolution and pose detection. Compared to existing products such as the Hailo-8 accelerator at 2-4 watts, the Ergo is targeting lower power (tens of milliwatts vs 2-4 watts for Hailo-8) albeit with lower performance.
Conclusions
As we’ve always said, the edge market is far easier to penetrate than training in the data center because different applications have dramatically different requirements. While an image processor for autonomous vehicles requires higher performance at higher power levels, a smart camera or an embedded speech to text processor demands a lower power envelope, and a lower cost. And many auto makers will prefer an SoC like NVIDIA Drive instead of designing their own SoC; Tesla is an exception to the rule.
Consequently, there is plenty of room for specific low-power processors such as Perceive, and the company is smart to quickly extend their first foray with a faster sub-watt processor. The software will be key to their success, enabling larger models to run efficiently on the new Ergo2.
We would note that there are many more competitors entering this market, so check back soon to stay up to date!
C-Suite News, Analysis And Advice For Top Decisionmakers
Sign up for the Forbes CxO Newsletter sent every Monday.
They can’t be too well informed
Getupthere
Regular
I think this article is a good sign that the leader in this space Akida is not mentioned at all.Was recently reading about Hailo 8. Uses 2 to 4 watts and they call that low power
market opportunity for 2023 is $5 billion dollars and growing.
Nice….
Only a matter of time folks.
DYOR.
Fact Finder
Top 20
Hi @Dhm
This article was posted the other day. It is a load of old string pulled together to make what appears to be a rope. Ergo & Ergo2 are pre trained single use chips no incremental on chip or one shot learning. You cannot add another type of processing once you have bought and deployed it.
As for the competitors they have chosen anything that sounds like it might have something to do with Ai.
Have a look at DeepAi. Don’t waste your time on it not worth the effort. A paid advertisement:
Now back to work.
My opinion only DYOR
FF
AKIDA BALLISTA
Similar threads
