750-P5-G5-D5-HI-A20-R When collecting edge data, edge computing devices usually need to connect 4K 120-frame high-resolution video cameras to display the acquired original image data smoothly and in real time. At the same time, the original image data is often a huge amount of data, which requires rapid and timely pre-analysis and data processing, so multiple acquisition cards or graphics cards are needed to meet the performance of the equipment, and with the help of high-performance graphics cards or even professional AI-level graphics cards, the acquired image data will be real-time AI processing and image rendering. However, the environment of edge devices is characterized by many uncertain factors and harsh conditions, which requires the quality of edge computing devices. In actual data acquisition scenarios, edge computing devices must be stable and reliable to avoid failures. Based on this, Advantech has designed and developed the EPC-B5000 edge computing device.
Rich network port expansion, help image data acquisition
In the edge-side AI machine vision image acquisition work, multiple high-definition cameras are needed, even 4K high-resolution, high-frame cameras are used for original image acquisition. At present, the devices on the market usually only have a single gigabit network port, which is difficult to support the collection of a750-P5-G5-D5-HI-A20-R large number of image and video data, so the demand for portable devices with multi-gigabit network ports is becoming more and more urgent. The EPC-B5000 has dual gigabit network ports and dual Gigabit network ports, and supports the connection of multiple cameras, which can effectively collect massive image and video data.
Portable and easy to use, the AI computing power can reach 50TFlops
Although the current portable devices are small in size, in AI computing, the physical space cannot support high-end graphics cards to do AI reasoning and calculation, so the edge computing devices need to meet the AI processing and be easy to move, plug and play. EPC-B5000 is a small portable edge device, which can also expand 2 professional-level AI expansion cards on the basis of space saving, improve AI computing power, reduce cloud processing pressure, distributed deployment, and reach 50TFlops computing power.
Stable operation in harsh environments, with a long life cycle
In the process of raw data acquisition and AI computing processing of edge computing devices, or even when the local data is transmitted to the cloud, the device is required to have a zero failure rate, which means that edge computing devices need to have strict quality, stable operation and no error. Generally, the edge side comp750-P5-G5-D5-HI-A20-R uting equipment is used in a harsh environment, with interference factors such as vibration, dust, and high/low temperature, and it is difficult to replace the equipment, and the cost of labor is expensive, so it is required to resist the harsh environment and have a long life cycle. EPC-B5000 supports wide temperature and passes China ESD level 4.IEC 62368. Certification can help operation and maintenance personnel greatly reduce operation and maintenance costs. At the same time, the EPC-B5000 has a life cycle of up to 15 years and the quality is stable and reliable.
Product characteristics
Support Intel® 10th-generation Core ™ i/Xeon W processors and Q470E/W480E/H420E chipsets;
4 x U-DIMMs, up to 128GB DDR4 2400/2666/2933MHz SDRAM;
Support triple display: 2 x DP++/VGA/eDP;
Supports PCIe Gen3, 16 x USB, 8 x SATA III, 2 x Gigabit port, 2 x 10 Gigabit port, 1 x M.2 M key;
Support dual BIOS function (Fsku);
4 x PCIe expansion slots.
Standard 4U height, support shelf
700W/1200W high power supply;
Support DeviceOn and embedded software APIs for large-scale and rapid AI deployment;
Support vPro, AMT 12.0 and software RAID 0.1.5.10, TPM 1.2/2.0 (optional).
Application case
In the new energy industry, EPC-B5000 can be used as a hardware computing control device at the bottom of the automatic power station to control the charge and discharge of the battery. It can be installed with two high-end AI computing cards, intelligent control of battery charging and discharging, high redundancy. By connecting to a high-definition display, the EPC-B5000 helps managers monitor battery status in real time, anticipate risks, and alert managers in time to avoid errors. It can also connect multiple sensors, after AI analysis, accurate and intelligent management of each sensor, a set of equipment can complete the entire automatic power station calculation control work.
