The increasingly important role that smart devices play in our lives is creating greater demand for these devices to deliver smooth and reliable experiences. It has been more than a decade since smartphones have become a common part of our everyday existence, but performance optimisation is still one of the biggest challenges facing the industry.
Despite the exponential increase of RAM and ROM on smartphones over the past ten years, slow app launching and switching speeds, phone overheating, and high-power consumption are still regular concerns for users. This leads to the question: If today’s smartphone hardware is so powerful, why do we still experience such fundamental issues in our user experience?
Powerful hardware only works if used right
Over the past decade, fierce competition in the smartphone industry has led to many manufacturers competing to deliver the most powerful device in terms of hardware performance. This has caused many industry analysts and everyday users to wonder if smartphone hardware has now become more powerful than it needs to be. However, the issue here is whether this powerful hardware is being used correctly.
When it comes to using powerful hardware in the most effective way, there are two common mistakes. The first is to allocate excessive computing resources to a process that doesn’t need them, leading to unnecessary power consumption. For example, two smartphones equipped with the same chipset can have vastly different levels of power consumption, and this is partly a result of the way resources have been allocated and scheduled.
The second type of misuse occurs during memory scheduling conflicts. These are what cause the phone to lag or freeze, for example when switching between applications, as system resources committed to the first app take time to become available for use by the second.
To solve these problems and ensure the efficiency of the hardware on its latest smart devices, OPPO has developed a system-level computing hub — the Dynamic Computing Engine — that can be found in the latest version of its mobile operating system, ColorOS 13.
The Dynamic Computing Engine: OPPO’s system level computing hub for ColorOS
ColorOS’s Dynamic Computing Engine uses a Computing Power Model to schedule hardware resources effectively and accurately. By optimising four computing technologies (Parallel Computing, High-Performance Computing, Device-Cloud Collaborative Computing, and Intelligent Computing), the Dynamic Computing Engine delivers a smoother and more stable user experience on OPPO devices in the long-term.
The Dynamic Computing Engine is an evolving technology platform that combines hardware and software to deliver maximum performance and efficiency. For the platform’s first edition on ColorOS 13, the Computing Power Model and the Parallel Computing Engine have been implemented to help solve performance issues related to inefficient resource allocation and memory scheduling conflicts.
The Computing Power Model: Striking a balance between high performance and low power consumption
Existing strategies for scheduling computing resources often lack a deep understanding of chip microarchitecture, failing to achieve the best balance in energy efficiency.
To solve these problems, ColorOS engineers underwent a three-year process to find the best way to allocate computing power and achieve an optimal balance between high performance and low power consumption. Following the analysis and simulation of such scenario data, the team developed a Computing Power Model capable of accurately scheduling CPU, GPU, and DDR computing resources at the instruction level.
With the support of the Computing Power Model, whenever the phone needs to render a game environment or open a camera application, the CPU only needs to perform the role of an “assistant”, rather than dedicating full focus to this one task. This avoids excessive resource requests and the need to repeatedly test resource allocation efficiency, thus reducing power consumption and extending battery life.
A look at data from the OPPO Find X5 Pro shows how the Computing Power Model improves power efficiency. According to data from OPPO Lab, once upgraded to ColorOS 13, a Find X5 Pro can be used to communicate via WhatsApp for 19 hours, play PUBG for 9 hours, view YouTube videos for 22 hours, or take WhatsApp video calls for 8 hours. While these are unlikely scenarios for the everyday person, simulations also show that the average battery life of the Find X5 Pro can reach up to 34 hours during typical use.
Parallel Computing: ensuring a smooth experience under heavy loads and multitasking
Just as instruction level resource scheduling strikes a balance between performance and power consumption, parallel computing is a valuable tool in solving memory allocation conflicts.
Android uses a serial memory allocation system that works on a first come, first served basis. However, given the complexity of real-life smartphone applications, this serial nature has significant drawbacks. For example, when the user suddenly needs to perform another high priority task, they may find that the phone’s memory is occupied by other processes and cannot be released in time.
This resulting lag or freezing of the app is an all too familiar experience for most smartphone users, and some wonder why such a common problem has yet to be resolved. One of the primary reasons is that these memory locks are caused by a phenomenon in the computer architecture called “lock contention”. In fact, research undertaken on ColorOS has shown that “lock contention” is responsible for 30% of the lag or freezing issues encountered when using a smartphone.
To get around this, ColorOS adopts an advanced microkernel design that executes key operations such as memory allocation and memory reclaim concurrently so that high priority tasks do not have to wait in line to have access to these resources.
By splitting the locked memory into smaller blocks, CPU threads have reduced wait time while the current thread finishes working on that specific block of memory, resulting in less lag time. This improvement is critical to the smooth experience that ColorOS users will enjoy during multitasking, enabling them to open multiple apps at the same time and switch between them.
A smoother experience in ColorOS 13
Thanks to the Dynamic Computing Engine, ColorOS 13 delivers a whole variety of smoother and smarter experiences for users.
When it comes to productivity, ColorOS 13 features an upgraded Multi-Screen Connect that now supports seamless connection between smartphones and OPPO Pad Air, and between smartphones and PCs.
With the OPPO Share function, the upgraded Multi- Screen Connect also makes file transfer between devices faster and more convenient, providing support for transfer between OPPO smartphones and OPPO tablets, or OPPO smartphones to PC, without using any mobile data. Most major file types are supported, meaning you can seamlessly and quickly transfer documents and switch between smartphone, PC, or OPPO tablet depending on which device works best for any given task.
In terms of interactive experience, ColorOS’s Quantum Animation Engine improves animated effects by adding extra details and an advanced physical motion model that makes interactions in ColorOS more realistic. Following in-depth research and analysis of user feedback, the ColorOS team has optimised 61 motion effects and introduced Behavioural Prediction into Quantum Animation Engine.
Behavioural Prediction can recognise and predict a user’s intended input when two touchscreen gestures are made one after the other. For example, when an app is opened and the user decides to open the second page to the right of the home screen, usually the user will have to slide up from the bottom of the display and wait for the home screen to appear before swiping onto the second page. Now, the user may swipe right after sliding up and ColorOS 13 will take the user directly there. This prevents conflict between the two operations and provides users with a smoother, more intuitive interaction.
Another feature in ColorOS 13 that minimises power consumption through close integration of hardware and software is LTPO 2.0 technology. LTPO 2.0 can reduce the refresh rate of the Always-On Display to only 1Hz, leading to a reduction in power consumption of up to 30% in certain applications.
This creates more possibilities for users to personalise their Always-On Display without having to worry about how it will impact battery life. The Always-On Display in ColorOS 13 provides options to monitor daily smartphone usage with Insight Always-On Display or create a digital Bitmoji avatar that syncs up with their real self. The Smart Always-On Display can even be used to control Spotify music or receive information on food delivery orders from apps such as Swiggy and Zomato without unlocking the screen.
In terms of privacy and security, ColorOS 13’s Auto Pixelate can automatically blur profile pictures and usernames in chat screenshots using a device-side algorithm, allowing users to conveniently share screenshots of their conversations without worrying about their privacy.
ColorOS 13 also includes an updated Private Safe feature that employs industry-standard AES file encryption to store files in a local private directory for enhanced security and privacy.
OPPO will continue to evolve the Dynamic Computing Engine in future versions of ColorOS to provide greater support for other features and content. This includes the addition of the High-Performance Computing, Device-Cloud Collaborative Computing, and Intelligent Computing engines that will take user experience to the next level.
This featured story is provided by OPPO.