Clarence Reynolds, TelecomTV (00:03):
The explosion of Edge AI and private 5G networks is transforming industries across the globe. Joining us today to discuss the challenges and innovations driving this revolution are Harini Malik Advanced Micro Devices Global Head Telco business at A MD and Gail Levin product marketing lead for HPE. Thank you both for being with us. Gail, I'll start with you. What are your clients using private 5G for?

Gayle Levin, HPE (00:31):
We've seen a big shift in what customers are using private 5G for, they're using it for the traditional use cases that we've seen like oil and gas to ensure worker safety airports to make sure that the luggage gets where you want it to go. And now we're starting to see it in other venues, especially large outdoor venues like stadiums where we see a combination of wifi for guests and we see private 5G for some of those critical services, like making sure that people get through the door and video analytics to help understand if the lines are getting too long so that they can streamline operations and increase their sales.

Clarence Reynolds, TelecomTV (01:06):
What key challenges are our organization facing in implementing Edge AI and private 5G solutions to enhance their operations?

Gayle Levin, HPE (01:14):
On the private 5G side, the challenge that we're seeing is there's been a lot of complexity. Traditionally, there have been a lot of moving pieces that need to be integrated, and the expectation was that the people who are working on it really understood 3G PP in the cellular space. What we're trying to do is standardize it to make it much easier for everyone to implement and creating that full stack, which we do with our friends at A MD to make sure that we can provide an end 10 solution and really accelerate and make it easier to adopt private 5G

Clarence Reynolds, TelecomTV (01:47):
And Harini. How do AMD's CPUs address the unique computational demands of Edge AI and private 5G applications across various industries?

Harini Malik, AMD (01:58):
EMD has a strong history of execution and we have the highest performance products, especially when it comes to compute in this market. And we are currently on our fifth generation Epic and Epic. The Epic products actually have the best performance as well as energy efficiency, which is one of the key requirements when it comes to 5G and especially when you start looking at Edge. Some of the key requirements for Edge is performance, but also energy efficiency that these processes provide. And Edge has some unique requirements when it really comes down to being able to reduce that latency especially and having that data at the edge and having the processing at the edge and A MD CPUs along with of the well-designed DL 3 25 year that we are exhibiting is a great combination for the private 5G.

Clarence Reynolds, TelecomTV (02:55):
Can you explain how this solution simplifies the deployment and management of low latency AI workloads at the Edge?

Gayle Levin, HPE (03:03):
We're leveraging the technology from a MD in two ways. One is that we are placing our industry leading mobile core and delivering it as an appliance to make it easier for enterprises to adopt, and that's leveraging everything that Heini just talked about. On the other side, we're also leveraging the compute capabilities so that once we're transporting the data over private 5G for AI analytics, we can do inference at the edge using all of the great capabilities and energy savings, speed and power that you actually need for AI that's in the server over here.

Clarence Reynolds, TelecomTV (03:40):
How does this solution enhance video analytics?

Gayle Levin, HPE (03:43):
One of the great use cases for private 5G is video surveillance and security, and we're seeing a lot of adoption and interest in that area. And one of the things that Purina can tell us a little bit more is about how once we transmit that data to the compute and at the edge, how we leverage a MD for that.

Harini Malik, AMD (04:03):
Actually, the market for video analytics is almost 75 billion by 2030, so the data at the edge and the processing of the data at the edge is becoming really prevalent because you don't want to take that data back to the centralized location. So having a high performance product and a processor that A MD provides as well as having that power efficiency at the edge becomes very important for video analytics. And we do that very well together,

Gayle Levin, HPE (04:34):
Right? It's something that you can't wait and send it up to the cloud. The video analytic has to happen at the edge it very quickly, and that's what we do together.

Clarence Reynolds, TelecomTV (04:44):
Harini and Gail, thank you for being with us today and sharing your insights.

Gayle Levin, HPE (04:48):
Thank you. Our pleasure.