Qualcomm recently unveiled the Cloud AI 100 edge AI development kit and announced the shipment of inference accelerators to select worldwide customers. The motive behind the accessibility of these products is to allow anyone to reap its benefits of AI performance.
According to the blog post, these products consume very little power and deliver an extensive boost in performance for companies that are hyperscalar in nature and looking to elevate their applications in artificial intelligence.
John Kehrli, Senior Director, Product Management at Qualcomm Technologies stated, “Today marks an important milestone, as the Qualcomm Cloud AI 100 inference accelerator is sampling with many customers and is poised to redesign the future of the data centres, 5G edge appliances as well as 5G infrastructure.”
Last year in April, the leading wireless technology innovator announced the Qualcomm Cloud AI 100 chipset to meet the explosive demand for AI inference processing in the cloud. The Qualcomm Cloud AI 100 utilises advanced signal processing and power efficiency.
Behind The Development Kit & Accelerator
The Qualcomm Cloud AI 100 Edge AI Development Kit, which is a 5G edge box, includes everything that the developers and customers need to build their best end-to-end AI solutions.
The developer kit is claimed to create 5G edge boxes that will change the retail and manufacturing landscape, among other industries. For instance, considering manufacturing, where safety is a key concern, the Qualcomm Cloud AI 100-equipped edge box can identify numerous people and objects simultaneously and ensure that protective equipment is being worn at all times.
Besides the cloud AI inference accelerator, the development kit also harbours the Qualcomm Snapdragon 865 Modular Platform along with the Snapdragon X55 Modem-RF, taking full advantage of Qualcomm Technologies’ offerings for application processing and 5G connectivity. It also provides support for various popular libraries including Pytorch, Glow, Tensorflow, Keras, among others.
The Cloud AI 100 Edge Development Kit is claimed to deliver breakthrough AI and video processing performance by providing features like advanced low power and high performance to 5G connectivity. This means the kit will help to accelerate the adoption of edge applications by offering a complete system solution for AI processing for up to 24 simultaneous 1080p video streams along with 5G connectivity.
The Qualcomm Cloud AI 100 edge development kit is an optimised system solution that offers high performance AI inferencing, video processing and 5G connectivity. The kit is powered by three modular Qualcomm components:
- Qualcomm Cloud AI 100 that provides low-power and high-performance AI inference processing.
- Qualcomm Snapdragon 865 Modular Platform that provides application and video processing.
- Snapdragon X55 ModemRF System Module helps in leading 5G connectivity in a pre-certified module offering global carrier support.
According to Kehrli, as a standalone product, the Qualcomm Cloud AI Inference accelerator is available in three form factors, which are DM.2e, DM.2, and PCI-e/HHHL. These accelerators will provide the customer flexibility with server and appliance designs with considerations to performance (up to 400 TOPS) and power (up to 75W).
He added, “The accelerator is also powered by the Qualcomm AI Cores, which are the driving force why Qualcomm Technologies’ offers the highest performing AI inference product to date.” These devices will be placed in data centres or onsite at the network edge, which will make a profound impact in the field of personalised recommendations, pedestrian safety, transportation, and communications.
Watch The Video Below:
If you loved this story, do join our Telegram Community.
Also, you can write for us and be one of the 500+ experts who have contributed stories at AIM. Share your nominations here.
A Technical Journalist who loves writing about Machine Learning and Artificial Intelligence. A lover of music, writing and learning something out of the box. Contact: firstname.lastname@example.org