ByLark Rochon 2020-03-06 310
Dipti Vachani, Senior Vice President of Arm and General Manager of Automotive and Internet of things, said, "for AI to be ubiquitous, device manufacturers and developers must implement terminal machine learning capabilities on billions or even trillions of devices. With the addition of these new forces to our AI platform, terminal machine learning will become the new normal even on the smallest devices, so that the potential of AI can be realized safely and effectively in a wide range of applications that are changing our lives. "
With the development of the Internet of things and the development of AI and 5G, more terminal intelligence means that smaller and cost-sensitive devices will become smarter and more functional, and will have higher privacy and reliability because they are less dependent on the cloud and the network. Arm brings more secure intelligence to microprocessors through its new design, reducing chip and development costs for product manufacturers who want to effectively improve terminal digital signal processing ((DSP)) and machine learning capabilities ((ML)), while significantly speeding up the time-to-market of their products.
Cortex-M processors have become the best choice for developers' computing platforms, and Arm partners have shipped more than 50 billion Cortex-M-based chips for a variety of customer applications. The new Cortex-M55 is Arm's most powerful AI processor in history, and it is also the first processor based on Armv8.1-M architecture and built-in Arm Helium vector processing technology, which can greatly improve the performance of DSP and ML while saving power. Compared with previous generations of Cortex-M processors, Cortex-M55 's ML performance can be improved by up to 15 times, while DSP performance can also be improved by 5 times, and has a better energy consumption ratio.
In addition, customers can also use Arm Custom Instructions (Arm custom instruction sets) to extend the processor's capabilities to optimize specific workloads, which is a new feature of Cortex-M processors.
For more demanding ML systems, Cortex-M55 can be matched with Ethos-U55, which is Arm's first micro-neuron processor, (microNPU), compared with the existing Cortex-M processor. The combination of the two can improve the ML performance of the product by 480 times.
Ethos-U55 is highly configurable and aims to accelerate the ML reasoning capabilities of space-constrained embedded and Internet of things devices. Its advanced compression technology can significantly save power and reduce the size of the ML model so that neural network operations can only be performed on larger systems before operation.
Arm is well aware that developer experience is critical to driving the AI revolution. As a result, Cortex-M55 and Ethos-U55 are fully supported by arm's industry-leading Cortex-M software development tool chain. In this way, we have a consistent development process for traditional workloads such as DSP and ML; at the same time, starting from TensorFlow Lite Micro, we carry out specific integration and optimization for advanced machine learning frameworks to ensure that developers have a seamless development experience and can get the best performance in any configuration of Cortex-M and Ethos-U55.
Arm believes that security should not be an ex post facto remedy, it is extremely important for the popularity of the Internet of things. To ensure the safest design and smooth PSA certification, these processors and paired Corstone reference designs can work with Arm TrustZone to ensure that security can be more easily integrated into a complete system-on-chip.
The new Cortex-M CPU and Ethos microNPU highlight Arm's commitment to promoting animal networking: to help chip designers and device manufacturers innovate with the Arm architecture even on the smallest end devices, and ignite a new wave of creativity and innovation for the Internet of things. This technology is widely supported by ecosystem partners and the industry, including Amazon (Amazon), Alif Semiconductor (Alif Semiconductor), Hengxuan Technology (Bestechnic), Cypress (Cypress), Dolby (Dolby), Google, NXP Semiconductor (NXP), Samsung (Samsung), STMicroelectronics (STMicroelectronics) and so on.
|You may also want to read:|
|Xiaomi Mi Watch Color VS HUAWEI Watch GT2 VS Amazfit GTR: Which is More Worth Buying?|
|Samsung Galaxy Buds+ VS AirPods Pro: What's the Difference Between These Two TWS Earbuds|
|How to Turn your Xiaomi Mi Band 4 as a remote shutter for Android camera?|
|Amazfit PowerBuds VS AirPods Pro: How to Choose the Most Suitable TWS Earbuds?|
|HUAWEI Honor Magic Earbuds (FlyPods 3) Review: Noise Cancellation Earphone with Low Budget|
|How to Unlock Your Windows PC / Laptop with Xiaomi Mi Band 4?|
● Over 300,000 products
● 20 different categories
● 15 local warehosues
● Multiple top brands
● Global payment options: Visa, MasterCard, American Express
● PayPal, Western Union and bank transfer are accepted
● Boleto Bancario via Ebanx (for Brazil)
● Unregistered air mail
● Registered air mail
● Priority line
● Expedited shipping
● 45 day money back guarantee
● 365 day free repair warranty
● 7 day Dead on Arrival guarantee (DOA)