As we know, the content our smartphones need to handle is becoming increasingly complex. Gone are the days of the first simple, monochrome displays; even today’s mainstream and lower cost devices now need to support a rich, multi-layered user interface and many of the latest applications and technologies. These complex use cases are pushing the boundaries of the requirements for Mainstream mobile, and with millions, if not billions, of users globally, it’s vital to be able to support them.
With this in mind we’re proud to announce the release of two new Mali GPUs, Mali-G52 and Mali-G31, to our Mainstream and Ultra Efficient roadmaps. Mali-G52 brings a brand-new execution engine design to provide greater performance in a smaller area, supporting premium use cases like Machine Learning (ML) and Augmented Reality (AR) in Mainstream devices, and Mali-G31 is our smallest GPU to date to support latest APIs for graphics development and complex User Interfaces for DTV. So how do they enable these next-gen technologies?
ML may seem like something you can only expect to see in the most sophisticated of mobile devices but you may be surprised to discover that it’s already happening more than you think. Even on mainstream phones, we expect our camera to be able to automatically detect our faces in order to focus accurately, or to be able to search our camera gallery for pictures of our cat, but you may not have realised this is facilitated by ML. Arm recognises the importance of ML and recently launched Project Trillium to address it. However, whilst our exciting new ML Processor is fantastic for premium ML workloads, it’s just one option in a much broader story. As we recently discussed, in mainstream and cost sensitive devices it often simply wouldn’t be viable to dedicate sufficient silicon to such a processor, but for smaller workloads like those I’ve mentioned, the GPU or CPU is often the right choice. We’ve talked before about how well suited the GPU is to this type of workload and also the ML-specific gains achieved by the latest generation of Arm Cortex CPUs.
What the newly announced Mali-G52 brings to the table takes things up a notch all over again. Building on the success of previous Bifrost based GPUs, Mali-G52 features a totally redesigned execution engine, the part of the processor which handles the arithmetic, to bring the higher performance needed for ML, within a small silicon budget. Mali-G52 is equipped with twice the number of lanes per execution engine, yet is only 22% bigger than its predecessor, allowing double the compute performance for all that complex content, without forcing our partners into big-budget silicon. Not only that, but Mali-G52 also introduces Int8 dot product support. ML inference on device makes extensive use of general matrix multipliers but you often don’t need FP16/FP32 level of precision. In many cases, Int8 is just as effective and much more efficient, and with the ability to handle four cycles per lane of each execution engine, Mali-G52 achieves nearly 4 times the ML performance of its predecessor on current image detection and other ML benchmark tests.
Our partners, too, appreciate how important this is. Daniel Diao, Deputy General Manager of the Turing processor business unit at Huawei, said:
ML is a top priority for us across all of our devices as user demands increase, so the ability to perform challenging ML workloads at such high quality on the GPU can make all the difference in the Mainstream market.
Yanqiu Daniel Diao, Deputy General Manager of the Turing processor business unit at Huawei
More immersive content too, is no longer solely the domain of the highest performing devices. Whilst you wouldn’t want to spend an hour with a full VR headset powered by the more efficient devices, (chances are you wouldn’t want to with any device, the fully immersive form factor is still working through a few kinks), that doesn’t mean alternative realities need escape you. AR is fast becoming the alternative reality tech of choice, and it’s vital to be able to support this, and other forms like 360 video, in the Mainstream. This is where Mali-G52’s best ever energy efficiency, combined with the smart task allocation capabilities of DynamIQ CPU architecture, really come into their own. The combination recommended in this generation of Mainstream Mobile solution is a DynamIQ configuration of one Cortex-A75 and seven Cortex-A55s. These multiple smaller cores are able to simultaneously handle a multitude of less taxing tasks, enabling the big and powerful Cortex-A75 to focus on the areas in which its full performance is truly necessary. Add in the Mali-G52, which achieves 30% better performance density than its predecessor due to the advances we’ve already talked about, and highly complex content is suddenly achievable. Much like the flexibility of DynamIQ, Mali-G52 features hugely scalable implementation options to allow our partners to implement exactly the balance of performance and efficiency their device requires. Partners can choose two or three execution engines per core, and up to four cores. This additional level of scalability is a fantastic benefit to the more cost constrained device tiers as it allows an incredibly granular level of trade-offs for each partner’s priorities.
It used to be the case that if you spent below a certain threshold on your mobile, you wouldn’t even consider gaming as an option beyond a basic Snake type experience. Now though, with displays ever more impressive and visuals more compelling, we’re raising our expectations in this area too. We talked a lot last year about High Fidelity Mobile Gaming pushing the mobile gaming industry beyond the realms of console and desktop gaming in terms of revenue and uptake, and this extends to Mainstream and ultra efficient devices too. This is where our second brand new GPU, Mali-G31, steps up to the plate. Designed with developers in mind, Mali-G31 is Arm’s latest ultra-efficient GPU, and the first on the Bifrost architecture. Mali-G31 is the smallest processor to be able to support not only OpenGL ES 3.2, but the more recent Khronos API, Vulkan, too. So what does this actually mean? It means that the vast energy efficiencies and optimizations enabled by these APIs are accessible to developers across a whole new performance tier of content. With millions of Mainstream and entry level devices currently active in the market, app developers can now optimise their content to work beautifully across all tiers of Mali powered device. This brings not only quality improvements, but quantity ones too. More games and applications will be available across more devices, allowing more people to play them. This means developers can reach a bigger market and sustain the next phases of their app development, something which can only be good for every step of the mobile food chain from silicon partner to casual gamer.
With innovations this significant continuing to push the boundaries of what we and our partners develop, the Mainstream market is gaining ground on the premium device to bring awesome, next generation visuals and compelling, immersive content, to a whole new tier of consumer.
Learn more about the Mali-G52
Learn more about the Mali-G31