There is no better way to explain the pace of change in the world of computing than by drawing parallels between Aesop’s hare and tortoise fable, and the two popular computing devices: smartphone and PC.
In a little over a decade, the humble phone has transformed into a mini-computer, while the PC languished. Smartphones have vroomed past their older PC sibling in speed, performance, and in some cases, storage. Mobile phones have become ubiquitous, and the microprocessors powering them have become more sophisticated.
Qualcomm has dominated the smartphone chipset market. And until Apple entered the scene to make its own system of chip (SoC), Qualcomm was the leading mobile computing chipmaker. The company’s Snapdragon processors have consistently enhanced the power and performance of smartphones over the years. Now, the San Diego, California-based company is looking spruce up the PC.
Kedar Kondap, SVP and GM, Compute and Gaming at Qualcomm Technologies spoke to The Hindu’s John Xavier about the convergence of the phone and PC, the era of GPUs, the importance of milliseconds in gaming, 5G, and competition in chip making for PCs.
Could you explain your role as leader of computing at Qualcomm?
Kedar Kondap: I was part of the mobile product management for over a decade, and about a year ago, I took this role to lead compute. This is a very important market; we’re looking to expand beyond mobile. I’ve been at Intel for seven years, prior to joining Qualcomm, so I’m familiar with the PC market in general. There’s a lot of convergence that we want to drive between a phone and a PC. So, we firmly believe that there is a lot of innovation needed in a PC. In the last 10 years, there has been so much innovation in smartphones, right from displays to camera. We want to drive some of those experiences into the PC which has seen very limited innovation in the last many years. So, our goal is to drive convergence of the phone and the PC.
Are you looking to kind of replicate what’s happening within the handset on a PC? Or are you looking at a totally new paradigm?
KK: We think of it in the context of user experiences. We want to bring the same user experiences from the phone to a PC. Over the years, most smartphone consumer experiences have become seamless, from booking a cab to GPS-based navigation. Multiple GPS satellites triangulate to get the best possible GPS signal to the phone. Sensors to track running and walking have also become efficient. This has reduced the drain on battery. All this has happened because a lot of intelligence has been put into the smartphone. It understands when a person is actively using their phone in certain areas. These are some user experiences we’re focused on. When we drive an architecture, we drive it with a power-first approach. Even today, if you ask any consumer, what is most important thing - whether it’s a PC, or a phone - they will tell you it is the battery life. They want the best battery life. So, we’re focused on such experiences that just doesn’t exist in a PC today. We want to change.
We are now in the GPU era as graphic computing plays a big role. How are you approaching these times? And what do you see as the next big thing in computing?
KK: In the last 10 years, GPU changed the way many things were used in a phone, or even a PC, with better UI with denser layers on smartphone. Ray tracing is now starting to come in phones. You get better shadows, and better rendering. In the next 10 years, we believe that there’s a third and important metric, which is AI. This will be completed by a connected intelligent edge technology. That means a lot of the processing is going to happen on the PC. We want to make sure that applications can take advantage of running on Snapdragon, and they take advantage of the heterogeneous architecture that we bring into PCs.
In phones today, if you look at Qualcomm architecture, you have a CPU and a GPU. Beyond that we have AI, a neural processing unit which is strong in terms of the number of networks that it runs, and the number of different objects that it can recognise. Then you have a dedicated video, and audio core. And then, dedicated sensors, and security features are tightly integrated. So, for example, if you’re running a video call, we want to be able to run dedicated aspects of the video call to take advantage of the architecture like linking your camera to face unlock through a secure pipeline through our SoC. These are things we want to bring from the phone to the PC.
In phones today, if you look at Qualcomm architecture, you have a CPU and a GPU. Beyond that we have AI neural processing unit that are strong in terms of the number of networks that it runs, and the number of different objects that it can recognise. Then you have a dedicated video, and audio core. And then, dedicated sensors, and security features are tightly integrated. So, for example, if you’re running a video call, we want to be able to run dedicated aspects of the video call to take advantage of the architecture like linking your camera to face unlock through a secure pipeline through our SoC. These are things we want to bring from the phone to the PC.
So, how are you planning that convergence?
KK: Our first step is to make sure all the apps that are running on smartphones today run seamlessly on Snapdragon. Second, they’re taking advantage of the architecture that we have. Then, they will go beyond that and take advantage of our AI for compute.
Intel, AMD, and NVIDIA are also looking to get share of the PC market. How do you see competition in this space?
KK: We recognise that there is an incumbent, but at the same time there is scope for innovation in PCs. It just hasn’t happened in the last many years. So, once you start bringing app developers to build applications, porting will start happening. That is why, along with Microsoft, we announced the Windows dev kit Volterra. We’re giving it out to multiple developers because once they start taking advantage of this architecture, porting stuff to Snapdragon will happen. And consumers will see a massive difference.
(For insights on emerging themes at the intersection of technology, business and policy, subscribe to our tech newsletter Today’s Cache.)
How do you plan to ride the 5G wave?
KK: There are many use cases. 4G solved bandwidth latency, which has allowed us to stream movies so easily on our smartphones. Now with 5G, it will just expand even more. We are working with a lot of developers in terms of latency reduction. I think that is an angle where we will bring in multiple use cases to PCs.
Could you share your perspective on the gaming market? How are things shaping up?
KK: If you look at the phone, as a category, we have Snapdragon Elite Gaming. And the reason why we separated out Elite Gaming from the rest is because there are several features that we particularly put in, such as Variable Rate Shading (which allows mobile games to run faster, at higher resolutions and lower power, while still maintaining the highest visual fidelity). There are so many things we’ve already done in this space. In this segment, a millisecond lag is the difference between winning and losing, particularly in eSports leagues. And so, we spent a lot of time reducing latency, understanding players’ WiFi network, and within that network, how is the player on static IP. All these things drive differentiation in Elite Gaming.