Arm-based SoC and demo board are available to test Morello 

To test the Morello architecture, developed by Arm and the University of Cambridge, Arm has designed and developed an SoC and demonstrator board which contains the first example of the prototype architecture.  

The Morello programme has been a research initiative by a consortium led by Arm to design a new, inherently more secure, Arm-based computing platform. Arm has been collaborating with the University of Cambridge for several years on its CHERI Capability Hardware Enhanced RISC Instructions) architecture, which defines hardware capabilities that would provide a fundamentally more secure building block for software. 

The CHERI architectural extensions are designed to mitigate memory safety vulnerabilities, or software defects that are exploited by hackers to take control of a device or system – at a hardware level. CHERI augments pointers (the variables in computer code that reference where data is stored in memory) with limits as to how those references can be used, the address ranges that they can use to access and which functionality they can use to access.

These hardware capabilities are unique to the processor architecture. Once baked into silicon, they cannot be forged in software. Use of these capabilities in place of some or all the memory addresses can improve the spatial memory safety of software, particularly software written in C or C++ code.

These capabilities can also be used as a building block to allow the enforcement of much stronger temporal memory safety with potentially far lower overheads than current approaches to partitioning. Known as compartmentalisation, this process isolates different parts of critical code into individual ‘walled’ areas. Code operating within one compartment has no access to any other area; even if an attacker breaches one piece of the code or data, they are trapped within that one small zone.

These hardware capabilities will be fundamental in designing future devices that are resilient to memory corruption vulnerabilities and other forms of software-based exploitation, explained Arm.

The Morello prototype boards are ready for software developers and security specialists to start exploring the security advances possible with the Morello architecture.

The limited-edition boards are based on the Morello prototype architecture embedded into an Armv8.2-A processor (an adaptation of the Arm Neoverse N1 processor). The boards are being distributed to major stakeholders, such as Google and Microsoft, as well as to interested partners in industry and academia via the UKRI Digital Security by Design (DSbD) initiative to test the hypothesis of Morello and discover if this is a viable security architecture for businesses and consumers.

The Arm Morello research program aims to create a more secure hardware architecture for processors. Its architectural extensions are based on the CHERI protection model.

The Morello program aims to assess the viability of the prototype hardware SoC employing unique extensions to the conventional Arm hardware instruction set that improve device security. 

“Computers are incredibly useful but the price we pay for that utility is more and more exposure to security and privacy issues,” said Ben Laurie, principal engineer, Security, Google Research. “CHERI can allow for better, more cost-effective protection without reduced performance and Arm’s Morello prototype can help mitigate security issues showing the way to a better future for all computer users,” he said.

David Weston, director of Enterprise and OS Security at Microsoft, declared he is excited about the Morello project. “Memory safety exploits are one of the longest standing and most challenging problems in all of software security,” he said. “Using core silicon architecture to eliminate whole classes of security issues with minimal performance impact has the opportunity to be transformative with massive positive impact”.

The next two years will see the ecosystem testing, writing code and collaboratively providing critical feedback to determine whether any features will be used in future versions of the Arm architecture, said Arm. If the Morello prototype architecture performs as expected, it will be fundamental in future processor designs, protecting businesses, individuals and the devices of tomorrow.

http://www.arm.com

> Read More

Industrial edge inference server is ready for rapid deployment

Powered by Nvidia’s A2 Tensor core GPU, the HPC-6240+ASMB-622 industrial edge inference server has been designed for rapid deployment, management, and scaling of AI and inference workloads in the hybrid cloud, says Advantech.

The HPC-6240+ASMB-622, powered by the soon to be available Nvidia A2, has a depth of 20.5 inch (520mm) depth and multiple expansion slots. 

Enterprises can confidently deploy hardware that securely and optimally runs accelerated workloads while using the Nvidia AI platform for inference, explained Advantech. Nvidia AI for inference includes software such as Nvidia Triton Inference Server and Nvidia TensorRT.

The server is therefore suitable for industrial equipment manufacturers (IEM), robotics, retail, intelligent video analytics (IVA), and other applications of AI at the edge.

The Nvidia A2 GPU’s compact size and low power requirements exceed the demands for edge deployments at scale, said Advantech. Combined with the HPC-6240+ASMB-622, it can deliver up to 20x higher inference performance versus CPUs and 1.3x more efficient IVA deployments than previous GPU generations, at an entry-level price point, added Advantech. 

The HPC-6240+ASMB-622 is a 2U, short-depth compact edge server with dual third Gen Intel Xeon scalable processors. It provides eight expansion slots and multiple PCIe for flexible GPU, network interface controller (NIC), frame grabber card and motion control card integrations. Four PCIe x16 design supports up to four Nvidia A2 GPUs for AI and high performance computing (HPC). 

The low power budget, compact size and multiple expansion capability with four A2 GPUs enables the server to be used for complex AI auto-optical inspection and manufacturing equipment applications.

The thermal management system increases airflow and pressure to enable high computing workloads at the industrial edge by cooling GPU cards and reducing noise output. 

The HPC-6240+ASMB-622 server is able to scale out from a single-GPU node to multi-GPU nodes when needed in industrial applications, especially for AOI, voice recognition and translation.

The HPC-6240+ASMB-622 is available now.  

Advantech specialises in IoT intelligent systems and embedded platforms. To embrace the trends of IoT, big data, and artificial intelligence, it promotes IoT hardware and software solutions with the Edge Intelligence WISE-PaaS core to assist business partners and clients in connecting industrial chains. 

http://www.advantech.eu 

> Read More

CES is a showcase for Gigabyte’s embedded design support

Edge servers, advanced driver assistance systems (ADAS) as well as motherboards for embedded systems available from Gigabyte, will be highlighted at this week’s CES in Las Vegas.

Its high-density edge servers are based on Arm architecture and Ampere’s 80 / 128- core CPU. They provide the performance, storage, network, and above all, the flexibility to be deployed at more challenging locations and limited spaces, says Gigabyte, which is increasing commonplace as processing work needs to be done on-site.

The speed of the 5G network will allow smart cities to use AI applications, for example transportation systems where edge computers can enhance the communications between devices to ensure traffic flows smoother. Gigabyte says its expertise for embedded computers allow it to have a clear vision of how to implement devices that can work in outdoor environments with stable functionality. These embedded computers come with GPU expandability to meet the needs of edge computing, and rich I/O interface designs to allow system developers to integrate multiple systems on a single platform, therefore developing applications such as licence plate recognition, real-time analytics, smart utility poles and smart traffic management.

Gigabyte also specialises in developing decision-making control units for different types of autonomous vehicles, as well as customised products such as ADAS engine control units (ECUs). The Gigabyte Pilot product series supports GPU and VPU acceleration modules, which combine various algorithms to provide precise object detection, decision-making and vehicle control capabilities. Products in the series are also equipped with high speed transmission interfaces which can be connected to a range of sensing devices to collect and transmit data in real time. 

Gigabyte and MyelinTek have developed a deep learning system called MLSteam, which includes optimised AI software stacking and comprehensive management tools such as resource, container and model management. MLSteam makes it easier for scientists to train and develop AI integration for autonomous vehicles and can make traffic management systems smarter, safer, and more collaborative, says Gigabyte.

The company’s latest embedded computing products feature a series of industrial motherboards and systems powered by Intel’s 11th generation processors. They are intended for factory automation, industry 4.0, smart retail, IoT and edge computing and are characterised by low power consumption. Examples of use are automated guided vehicles, which can be programmed to manoeuvre along designated routes, detect and avoid obstacles through navigation, and also industrial automation security systems. 

AI has also been integrated for visual recognition to upgrade digital signage to detect users’ gender, age to improve retail experiences. Restaurants and delivery services can use electronic control units for smart delivery services, using autonomous vehicles. AI and IoT can also be used to automate the operation and management of public spaces rentals for office, conference, dining and parties.

The Z690 and Aero motherboards will also be highlighted at CES. 

https://www.gigabyte.com

> Read More

NanoCOM-TGU combines 11th Gen Intel Core processor with UHD Graphics

A COM Express Type 10 module, powered by the 11th Generation Intel Core processor (Tiger Lake UP3), has been released by Aaeon. The NanoCOM-TGU optimises embedded and mobile applications for today’s IoT requirements, says Aaeon.

The NanoCOM-TGU’s computing capability is combined with the graphical processing power enabled by Intel UHD Graphics in a compact COM Express Type 10 mini form factor, measuring 84 x 55mm.

Power by the 11th Generation Intel Core SoC, the module is designed for applications with intensive data and graphic processing needs. The SoC integrates both CPU and GPU, as well as AI and deep learning acceleration engines, to accelerate processing capability in embedded mobile applications, such as telematics, public sectors, and industrial automation.

According to Aaeon, the mini form factor of COM Express Type 10 of NanoCOM-TGU shortens the design and deployment processes for system designers or integrators in embedded mobile projects. In addition, by leveraging the Type 10 form factor, NanoCOM-TGU offers the high-speed I/O interfaces to shorten the system design, including one 2.5Gigabit Ethernet, two display outputs (eDP and DDI), one high-definition audio output, two SATA 3.0 interfaces, one onboard PCIe NVMe SSD (up to 256Gbyte), 10 USBs (allocated as eight USB 2.0 and two USB 3.2 Gen 2 ports), two UARTs and one PCI-Express 4. 

For memory bandwidth and storage, NanoCOM-TGU supports LPDDR4x memory with in-band ECC up to 16Gbyte for IoT and big data related applications. The onboard PCIe NVMe SSD interface supports up to 256Gbyte and accelerates read / write speeds multiple times than conventional storage interfaces. 

Aaeon adds that system designers or integrators can use the Intel UHD Graphics’ graphical processing power for enhanced visual performance. 

Established in 1992, Aaeon designs and manufactures professional intelligent IoT products. It provides reliable and high quality computing platforms, including industrial motherboards and systems, industrial displays, rugged tablets, embedded controllers, network appliances and related accessories and integrated solutions. The company also has the hardware and services for premier OEM/ODMs and system integrators worldwide. 

As an Associate Member of the Intel Internet of Things Solutions Alliance, Aaeon offers customised services from initial product conceptualisation and board product development to mass manufacturing and after-sales service programs.

 

https://www.aaeon.com

> Read More

About Smart Cities

This news story is brought to you by smartcitieselectronics.com, the specialist site dedicated to delivering information about what’s new in the Smart City Electronics industry, with daily news updates, new products and industry news. To stay up-to-date, register to receive our weekly newsletters and keep yourself informed on the latest technology news and new products from around the globe. Simply click this link to register here: Smart Cities Registration