What you’ll learn:
- The right approach to building software-defined vehicles.
- Why hardware-enforced graphics virtualization technology is an industry game-changer.
- How SR-IOV compares to VirtIO in terms of performance and security.
Discussions around software-defined vehicles (SDVs) and how they will change the way carmakers design and build cars have been happening for over a decade. This necessary paradigm shift is driven by the need for automakers to meet the evolving expectations of consumers for more personalized and dynamic experiences.
Think of a higher frame rate during gameplay, 3D map applications instead of 2D, real-time 3D visualizations across multiple displays within the vehicle, or enhanced safety thanks to real-time AI inferencing. The key to unlocking these next-gen experiences that consumers crave is silicon-enforced virtualization.
Virtualization plays a part in the solution as more software components are combined on a single system-on-chip (SoC) with GPUs. Access to the GPUs via a hypervisor can be done using virtual-only separation (VirtIO) or single-root virtualization (SR-IOV)
Moving Beyond the Hypervisor Will Revolutionize In-Vehicle Compute
Until now, the auto industry has been trying to move toward the software-defined promise by using a hypervisor for software virtualization, creating a bottleneck that can’t scale with the performance demands of today’s workloads and certainly not the workloads we would expect from vehicles in the future.
For the potential of SDVs to truly be realized, they must be supported by hardware-enabled virtualization to reinforce the most efficient individual workload path. Intel offers the silicon-enforced separation that bypasses the hypervisor, adding performance within the software for higher quality and new workloads (Fig. 1).
Think for a moment about the computing needed to power an SDV as if it were an electric vehicle (EV) about to embark on a road trip with a fully charged battery. It’s generally accepted that if it leaves home (Point A) to go directly to its set destination (Point B), it optimizes performance—in this case, the vehicle’s range. That’s how Intel’s silicon-enforced virtualization works—it makes an efficient trip to the hardware.
However, if the EV is forced to make a detour to an alternative location (Point C), it must use up vital energy, and the trip takes longer. This forced “detour” is similar to the experience delivered by other silicon providers, which ultimately leads to significant performance degradation.
Performance Differences Between SR-IOV and VirtIO
SR-IOV provides a robust and high-performance, low-overhead, hardware-based virtualization solution that eliminates the risks associated with virtual-only separation, VirtIO. When using Intel SDV SoCs with SR-IOV, each workload is separated directly at the GPU silicon level, freeing up the software layers for enablement of additional performance and functionality with zero added latency.
In terms of security, SR-IOV is also more equipped to defend against risks because it leverages hardware virtualization capabilities. While VirtIO presents a larger attack surface within the virtualization layer (between the guest OS and host kernel/virtual switch) that can potentially compromise the security of the service OS, Intel Graphics SR-IOV doesn’t contain such software virtualization layers, therefore minimizing the attack surfaces for interference and denial-of-service exploits.
Virtualization Solution Outperforms Competition
By leveraging the company’s silicon-enforced virtualization capabilities, Intel Automotive can offer the industry a more performant and efficient approach to designing a software-defined vehicle (SDV)—one that delivers about 99% efficiency and zero added latency.
Case in point: Using the industry-standard graphics benchmark, GFX Manhattan 3.0, Intel-based GPU SR-IOV was able to operate a single workload at 99% efficiency compared to VirtIO, which ran at 43% efficiency (Fig. 2). Therefore, if you run a workload that needs 100 frames per second (fps) with the Intel solution, you get 99 fps with zero latency. With the alternative solution, you get 43 fps plus additional workload-dependent latency.
GPU Tech and SR-IOV Virtualization Redefine In-Vehicle Infotainment
Automakers have made it clear that the demand for graphics performance stands out as the primary area of growth when considering requirements for an in-vehicle infotainment system. When implementing these infotainment systems, virtualization is a key technology to achieve multifunctionality.
Crucially, the choice of graphics virtualization technology can significantly impact the overall performance, security, and stability of the automakers' deployed solution. Without the appropriate virtualization features in silicon, the advertised capabilities of the chosen system will not be met.
