This Week in Analytics & AI News

One of the United States’ first exascale systems. A better way to share patient data. Supporting the rise of AI in India. A global health and sciences collaboration. And the benefits of local data processing.

HPC Workload Convergence Paves the Way for AI in the Exascale Era

HPC is from Mars and AI is from Venus. The two make for one heck of a power couple when they converge, but different workflow requirements can put them at odds. To smooth things over, Intel and Argonne are teaming up to deploy an Exascale system that can deliver across a range of workloads.

Intel Works With Penn Medicine in Using Privacy-Preserving AI to Identify Brain Tumors

Sharing is caring, especially when it comes to information that trains AI models to assist the early detection of brain tumors. But the exchange of patient data is fraught with legal complications. So researchers are turning to a distributed machine learning technique that protects patient privacy.

How Intel Is Helping Accelerate the Uptake of AI in India

The proliferation of AI across India has made AI specialist one of the most sought after roles in the country. But for all the enthusiasm, leadership has been hard to come by. Leveraging its long-standing relationship with the Indian government, Intel programs are giving the country’s AI initiatives legs.

Technology in a Time of Crisis: How DARPA and AI Are Shaping the Future

Unprecedented times call for unprecedented solutions. In addition to the widespread collaboration that’s taking place among the global health and sciences community, emerging technologies like AI and advanced analytics are helping identify potential treatments, automate screening, and care for patients.

3 Key Benefits of Training at the Network Edge With AI

The number of devices that connect to the internet is rising fast. And with more IoT comes a lot more data. But storage for these devices and their associated data is expensive. Rather than send potentially useless data to the cloud, why not keep that data local by processing it at the edge?