BRN Discussion Ongoing

[Fullmoonfever]

We are manipulated by shorters, Big Named deals and Revenue is the only hope

Not true....we could try reach out to Ben Kenobi

 

[Iseki]

Hi Manny,

I have in the back of mind that there is a physical Akida 2 FPGA available to customers for testing (stage 2 testing after having fine tuned your model on the cloud FPGA). Once you make one, you just use a cookie cutter to reproduce on demand.

Hi Manny,

I have in the back of mind that there is a physical Akida 2 FPGA available to customers for testing (stage 2 testing after having fine tuned your model on the cloud FPGA). Once you make one, you just use a cookie cutter to reproduce on demand.

Nah, But BRN will have a version that will run on Xilinx or Frontgrades FPGA chip.

I wonder what IP EDGE-AI will use for the other CPU chip. SiFive? Arm? Hopefully RISC-V.

On a good note Korea's well funded start-up Rebellions looks like it's targeting NVIDIA, no?

 

[Newk R]

Any more announcements and we'll go broke

 

[Bravo]

EDGEAI lists MDS Intelligence as a client on its website!


1.

On its own website, MDS Intelligence lists RapidMetering as one of its core digital-transformation products.

The product is described as a remote water-metering solution designed to improve environments where analog meters are still in use, ntended to reduce or eliminate manual meter reading.

Technically, the system works by:

• capturing images of analog meter dials using a camera
• transmitting those images to a server
• using AI / deep-learning analysis to convert the images into digital meter readings.

The system reportedly achieves over 99.5% recognition accuracy when interpreting meter readings.

An important detail is that RapidMetering today does not appear to rely on a custom AI chip. The AI processing seems to occur off the device, as the article describes the sequence as:

• camera → transmit image → AI analysis → digital reading

If the architecture were edge AI, the sequence would normally look more like:

• camera → local AI inference → send meter reading

I suppose moving the AI inference to the edge would offer several advantages inlcuding lower bandwidth, faster response times and reduced cloud compute costs.

For that reason, I think it is quite plausible that EDGEAI could be developing a chip intended to move this type of AI inference from the cloud to the edge.

The name “RapidMetering” used by MDS and the phrase “Rapid Meter” in the BrainChip announcement may therefore not be a coincidence.
Nor might the reference to an 8-year battery life, which appears consistent with the durability mentioned for the RapidMetering system.







2. Extract from MDS article dated 10 Jan 2025 (full article linked below)

3.

4.




EDGEAI Website

엣지에이아이(주)

엣지에이아이(주)

www.edge-ai.kr

MDS Intelligence article dated 10 Jan 2025

MDS인텔리전스, CES 2025서 수도 원격 검침 기술 선보여

MDS인텔리전스가 CES 2025에서 AIoT 기술을 활용한 수도 원격 검침 솔루션 ‘래피드미터링’을 공개하며 스마트 관망관리 기술의 우수성을 세계에 알렸다.

www.venturesquare.net

 

[jtardif999]

All good. Glad there are no confusing price metrics attached to this announcement. A positive milestone for ADK2000.

Nothing to do with AKD2500 (the chip BrainChip is fabricating); this is an IP deal regarding the Akida2.0 RTL being laid down as IP in EDGEAI chips .