editor's blog
Subscribe Now

CEVA Goes for Base Stations

Mobile communications have been one of CEVA’s focus areas (others being audio and images). If you’re new to CEVA, they do DSP cores for SoCs, focusing on low power as a critical feature. (They have lots of hardware features, but at the end of the day, whether it’s a hardware accelerator or an optimized instruction set, it all leads to lower power and longer battery life.)

We’ve covered them before (albeit getting distracted by the incredible alphabet soup that characterizes this market). As complexity has grown, they’ve seen the need for multiple DSP cores, so they put together a multicore platform.

But most of their mobile effort was going into DSPs that would reside in a handset. And yes, handsets have being going multicore for lots of reasons. And with the proliferation of smartphones, they have to be the most abundant example of heterogeneous multicore. In other words, different cores for different purposes – applications, baseband, graphics, etc. This requires an asymmetric model, with every core having its own OS and memory image (possibly sharing some memory for message passing and such).

But now they’re going for more than just the handset: they’ve just introduced a new XC4500 family that focuses on mobile infrastructure – and, specifically, base stations. You might think this would just be a bigger version of what they use in the handset, which is the XC4000 family. But it’s not, because what happens in a base station is very different from what happens in a phone.

A handset is all about taking a single call or session or whatever and breaking it down to extract the content and send that content to the appropriate places in the phone. That’s not at all what a base station does; it manages traffic. It doesn’t care, for the most part, what’s happening with any particular call or session; it’s just making sure everything gets to the right place. This is, basically, packet processing.

So while the phone needs all these different processors to handle the different aspects of the content, the base station simply needs to be able to scale what it does to accommodate the amount of traffic it has to handle. Which means that, unlike the phone, it can benefit from a homogeneous multicore architecture using a symmetric approach (SMP). If one core can process x calls, then n cores can process n*x calls. More or less (yeah, I know it’s not quite that simple…).

Which makes the XC4500 look different from the XC4000, even though they’re on opposite ends of the same airwave. It’s much more like a router than it is like a phone. Because it is a router of sorts. Traffic management features allow multiple independent queues and provide built-in dynamic scheduling. Data for a specific task is stored in shared memory, so assigning it to a specific core merely involves sending a pointer rather than a time-consuming data copy. They have cache coherency infrastructure to keep all of the cores’ caches in synch as well.

You might wonder, by the way, what the opportunity is for new base stations. And, apparently, there’s not a lot of movement in the traditional fiber/cable-backhaul market, where your wireless call gets sent to the mothership over a wire. But new installations are starting to favor wireless backhaul over microwaves. That’s where they see things looking up.

You can find out more in their release.

Leave a Reply

featured blogs
Dec 2, 2024
The Wi-SUN Smart City Living Lab Challenge names the winners with Farmer's Voice, a voice command app for agriculture use, taking first place. Read the blog....
Dec 3, 2024
I've just seen something that is totally droolworthy, which may explain why I'm currently drooling all over my keyboard....

featured video

Introducing FPGAi ā€“ Innovations Unlocked by AI-enabled FPGAs

Sponsored by Intel

Altera Innovators Day presentation by Ilya Ganusov showing the advantages of FPGAs for implementing AI-based Systems. See additional videos on AI and other Altera Innovators Day in Alteraā€™s YouTube channel playlists.

Learn more about FPGAs for Artificial Intelligence here

featured paper

Quantized Neural Networks for FPGA Inference

Sponsored by Intel

Implementing a low precision network in FPGA hardware for efficient inferencing provides numerous advantages when it comes to meeting demanding specifications. The increased flexibility allows optimization of throughput, overall power consumption, resource usage, device size, TOPs/watt, and deterministic latency. These are important benefits where scaling and efficiency are inherent requirements of the application.

Click to read more

featured chalk talk

High Power Charging Inlets
All major truck and bus OEMs will be launching electric vehicle platforms within the next few years and in order to keep pace with on-highway and off-highway EV innovation, our charging inlets must also provide the voltage, current and charging requirements needed for these vehicles. In this episode of Chalk Talk, Amelia Dalton and Drew Reetz from TE Connectivity investigate charging inlet design considerations for the next generation of industrial and commercial transportation, the differences between AC only charging and fast charge and high power charging inlets, and the benefits that TE Connectivityā€™s ICT high power charging inlets bring to these kinds of designs.
Aug 30, 2024
36,118 views