Congatec COM HPC Client Modules

2022: Release of conga-HPC/ cALP COM-HPC Client Modules with 12th Generation Intel Core Processors

On January 4, 2022, congatec, a supplier of technology for embedded and peripheral computing, introduced 12th generation IntelCore processors for mobile and desktop PCs (formerly code name Alder Lake) on 10 computers-on-modules COM-HPC Client and COM Express. Featuring high-performance cores from Intel, COM-HPC A and C form factors and COM Express Type 6 form factors offer performance gains and improvements for embedded and peripheral computing.

Engineers can now use Intel hybrid architecture with high performance. Offering up to 14 cores/20 threads in the case variant with BGA ball pins and 16 cores/24 threads on desktop versions of LGA (a variant of the processor package using a contact pad array located on the case), the 12th generation Intel Core processors provide a qualitative leap in multitasking and scalability for next-generation IoT and peripheral applications. Congatec modules with Intel processors receive up to 6 or 8 (BGA/LGA) optimized Performance (P-core) cores, as well as up to 8 energy-efficient efficient Efficient (E-core) cores with low native power consumption, as well as faster multi-threaded applications and more efficient background tasks support DDR5 memory.

Mobile processors running BGA with up to 96 execution modules of the Intel Iris Xe integrated graphics processor are processors11th Generation Intel Core estimated to provide up to 129% better graphics performance for immersive user interaction, and can handle parallel workloads such as. algorithms artificial intelligence (AI)

Optimized for maximum embedded client performance, the graphics of processor LGA-based modules now provide up to 94% performance, and the output performance of image classification (the process of extracting information from multi-channel raster images) has almost tripled with an increase in throughput of up to 181%. In addition, the modules offer wider bandwidth for connecting discrete graphics processors in order to achieve maximum performance in terms of graphics and AI performance based on GPGPU (the concept of using a graphics accelerator to perform calculations instead of a central processor). Compared to BGA versions, these and all other peripherals benefit from double the transfer rate, since they are equipped with a fast PCIe 5.0 interface in addition to PCIe 4.0 outside the processor. In addition, desktop chipsets provide up to 8 PCIe 3.0 lines for additional connectivity, and BGA mobile options offer 16 PCIe 4.0 lines off-processor and up to 8 PCIe 3.0 lines off-chipset.

Target industrial markets for BGA and LGA processor variants can be found wherever high-performance embedded and peripherals are deployed. This includes, for example, peripheral computers and IoT gateways (IoT), which include several virtual machines for intelligent factories and process automation, AI-based quality control and machine vision, real-time collaborative robotics and autonomous vehicles for logistics, warehouses and delivery. Typical off-site applications include autonomous vehicles and mobile machines, video surveillance applications and gateways in transport and smart cities, as well as 5G cloud devices and peripherals that require inspection of artificial intelligence packages.

Using Intel's innovative performance hybrid architecture delivers impressive P-core performance combined with energy-efficient E-cores. Intel Thread Director assigns each workload to the appropriate cores for optimal performance. Selected Intel processors with Time Coordinated Computing (Intel TCC) and Time-Sensitive Networking (TSN) are also suitable for real-time hard applications. Combined with full support for hypervisor technology from Real-Time Systems, they are a platform for consolidating many different workloads on a single edge platform. Since this applies to both low-power and high-performance scenarios, it allows you to create environmentally friendly structures with minimal environmental impact. explained Christian Eder, Marketing Director for congatec

In addition to high throughput and performance, the flagship COM-HPC Client and COM Express Type 6 modules feature dedicated artificial intelligence engines that support Windows ML, a high-performance and reliable API for deploying hardware-accelerated machine training-capable solutions on Windows devices, Intel Distribution of OpenVINO toolkit (an open, free toolkit that helps data developers and analysts accelerate the development of high-performance solutions for use in various video systems) and machine learning with Chrome Cross ML. Various AI workloads for processing intensive peripheral AI workloads can be easily delegated to P-cores, E-cores, and GPUs. Intel's built-in Deep Learning acceleration technology uses different cores using vector instructions of the VNNI (Vector Neural Network Instructions), and the built-in graphics support DP4a GPU instructions with AI acceleration that can be scaled even to dedicated graphics processors. In addition, Intel's least-power AI accelerator, the Intel Gaussian & Neural Accelerator 3.0 (Intel GNA 3.0), provides dynamic noise suppression and speech recognition, and can work for voice wake commands even when the processor is in a low power state.

Combining all of the above features with Real-Time Systems hypervisor technology, as well as operating system support for real-time Linux and Wind River VxWorks, makes congatec modules a truly comprehensive ecosystem package to facilitate and accelerate the development of applications for high-performance peripheral computing.

Compact conga-TC670 COM Express Type 6 modules based on the 12th generation Intel Core mobile processor (95 x 95 mm) and conga-HPC/ cALP COM-HPC Client modules of size A (120 x 95 mm) will be available in the following configurations:

conga-HPC/ cALS COM-HPC Client modules of size C (120 x 160 mm) based on 12th generation Intel Core desktop processors will be available in the following options:

Note: * Processor Base Power (PBP) - formerly called PL1 State, it is equal to the TDP value. PBP characterizes processor consumption at rated operating frequency.

2021: Release of Modules on the 11th Generation W-11000E and Intel Celeron processors Intel Core vPro , Intel Xeon

On August 5, 2021, Congatec, a supplier of embedded and peripheral computer technologies, introduced twenty computers-on-modules for the Internet of Things. These modules, featuring the 11th generation Intel Core vPro processor, Intel Xeon W-11000E processor, and Intel Celeron processor, are designed for the most performance-demanding peripheral computing applications and IoT gateways in terms of bandwidth.

The flagship client modules COM-HPC Client and COM Express Type 6 are based on Intel SuperFin 10nm technology in a dual-component package with a dedicated CPU and Platform Controller Hub (PCH).

Flagship applications that directly benefit from these GPU enhancements can be found in surgery applications medical visualization , and to telemedicine because for optimal diagnostics, congatec's updated platform supports video in 8K HDR resolution. Combined with platform capabilities artificial intelligence and a comprehensive set of Intel OpenVINO tools, physicians can easily access diagnostics and understand diagnostics data through deep learning. But this is just one of the qualities of Intel UHD built-in graphics, which also supports up to four displays 4K at a time. In addition, these modules, thanks to 11th generation Intel processors, can process and analyze up to 40 HD 1080p video streams at 30 frames per second in parallel for 360-degree circular viewing in all directions. These capabilities of mass, machine vision based on artificial intelligence, are also important for many other markets, including automation of production, machine vision for quality control in production, the formation of safe spaces and cities, as well as collaborative robotics and autonomous transport means, in,, and to logistics agriculture construction public, and this is transport far from all.

AI Logic Output and Deep Machine Learning algorithms can work seamlessly, either in mass parallel mode on an embedded graphics processor or on a processor, using built-in deep learning workload acceleration with Intel Deep Learning Boost, which combines three instructions into one, accelerating logic output processing and raising situational awareness.

The presented client platforms COM-HPC Client and COM Express Type 6 have built-in security features that are important for the failure-free operation of many mobile vehicles and robots, as well as stationary equipment. As for such applications the support in real time is obligatory, modules of the congatec company can start the operating system of real time (OSRV, English real-time operating system, RTOS), such as Real Time Linux and Wind River VxWorks and to provide the built-in support of technology of a hypervisor of the Real-Time Systems company which is also officially supported by Intel processors. The result for customers is a truly comprehensive ecosystem package with as much support as possible. Additional real-time capabilities for real-time IIoT/Industry 4.0 gateways and peripherals include Intel Time Coordinated Computing (Intel TCC) and Time Sensitive Networking (TSN). Advanced security features that help protect systems from attacks make platforms a good candidate for all types of mission-critical client applications in businesses and utilities.

The conga-HPC/ cTLH COM-HPC Client Size B (120 x 120 mm) modules and the conga-TS570 COM Express Basic Type 6 (125 mm x 95 mm) modules will be available with a scalable 11th generation Intel Core processor, featuring a selection of Xeon and Celeron processors and are designed even for extreme operating temperatures in the range of 85 + Both form factors support up to 128 GB of DDR4 SO-DIMM memory at 3200 million transactions per second and additional ECC - a type of computer memory that automatically recognizes and corrects spontaneous changes in memory bits. For high bandwidth peripherals, COM-HPC modules support 20 PCIe Gen 4 lines (x16 and x4) and COM Express versions support 16 PCIe lines. In addition, developers can use 20 PCIe Gen 3 lines with COM-HPC and 8 PCIe Gen 3 lines in COM Express.

The COM-HPC carrier module provides the 1x PCIe x4 interface to support the NVMe ultra-fast SSD. In addition, the COM Express board has a built-in NVMe solid-state drive for optimal use of all PCI Gen 4 lines supported by the updated processor. Additional media can be connected via 2x SATA Gen 3 on COM-HPC and 4x SATA on COM Express.

The COM-HPC module offers 2x USB 4.0, 2x USB 3.2 Gen 2 and 8x USB 2.0, and the COM Express module according to the PICMG specification offers 4x USB 3.2 Gen 2 and 8x USB 2.0. The COM-HPC module offers 2x 2.5 GbE for network operation, while the COM Express module has only 1x GbE, but both support Time Sensitive Networking (TSN). Audio in the COM-HPC version is provided through I2S and SoundWire interfaces, and in COM Express modules - through HDA. Comprehensive board support packages are available for all leading SRVs, including hypervisor support from real-time systems, as well as Linux, Windows, and Android.

Both modules based on the 11th generation Intel Core, Xeon and Celeron COM-HPC and COM Express Basic Type 6 processors are available in the following options:

2020: COM-HPC Client size A

On September 3, 2020, congatec, a provider of embedded technology, announced the release of its first two products based on them - a client module in form factor A - COM-HPC Client size A and a next-generation compact computer-on-module COM Express. More details here.