Arm Tech Symposia China: Security, Connection and the Complete Display Solution

It’s a glorious sunny day here in Shanghai and we’re gathered in the very beautiful ballroom of the Kerry Hotel. A wall of screens shows Arm’s latest thinking in revolutionary technology areas like Machine Learning, IoT and more as hundreds of people take their seats. Videos of the engineers and thinkers of the future play out how they see the next world-changing technologies in readiness for the first of today’s keynotes.

As the lights dim we’re treated to an insight into Arm’s vision for the immense range of possibilities in the coming age of technology. First up on stage is Allen Wu, EVP and VP of China, presenting his vision for working closely with China in building an ecosystem of open innovation and growth, from 1 trillion IoT devices, to Artificial Intelligence which exceeds that of humans, until we reach such a stage of connected intelligence that we’re actually able to predict the future. My Mandarin is a little rusty (Ok, non-existent) so I can’t tell you all the details of his announcements but I can assure you the collaboration between Arm and China is extremely well received by our hundreds of distinguished guests and sponsors as an opportunity to shape the China of tomorrow.

The future of Arm

Next up was Arm Chief Strategy Officer, Jason Zajac, who joined us to share the strategic direction of Arm’s future plans. Just into his second six months with Arm, Jason has hit the ground running and is excited to be able to meet with our partners at events such as this and learn what we can do to support their goals. A year or so on from when Arm joined the Softbank family, Jason is more than happy with how productive the collaboration is proving. He shared the vision, optimism and forward thinking that inspires him every time he meets with Softbank founder Masayoshi San and sees his big-thinking, world changing visions and goals. One of these goals is to deliver the next 100 billion Arm based chips in just the next 4 years, and there are a few specific things that will make that happen…

Panoramic shot of the presentation suite

Big data and AI are the topics everyone is talking about and Jason was intent on reassuring that whilst AI can inspire some nervousness, the positive benefits of these technologies far far outweigh the concerns. Our recent study on the subject actually showed that most people are incredibly positive about the future of AI, and this was never more evident than in Asia, where almost three quarters of those asked were excited about the future of AI and the benefits it can bring to our lives. There is such a positive embracing of technology across Asia already that it’s by far the most exciting market in terms of willingness to accept the uncertain. A human-like chat bot in social media platform Weibo is already incredibly popular and people are engaging with it in increasingly long and complex interactions. The fact that something like this could already be so lifelike as to inspire personal confidences in the dead of night, shows just how far this technology has come already, and just the beginning of the value it can add to our personal lives alone.

AI and Machine Learning technology is also already complementing and enhancing human capability in other areas, and not replacing us as some have feared. 47% of people would already use an AI doctor, as so much of diagnostics is information gathering and pattern recognition that it’s a field so naturally attuned to deep learning and AI that it can be hugely beneficial. It can detect skin cancer much faster, earlier and more reliably than human medicine which is hugely encouraging as our health systems become ever more stretched.

Autonomous driving is of course another HUGE topic of conversation and again, consumers are actually far more confident in the safety implications of autonomous vehicles than one might expect, especially when you consider that the industry is not even ready to deploy it yet. Of course, all of these smart technologies, and especially safety critical ones like autonomous driving, simply MUST be secure. Jason talked about the recent case of a fish tank which was intelligently temperature controlled. This might not seem like a huge focus area for security but unfortunately, the ability to access the controls of the fish tank became a back door into the overall system, allowing a breach through an area that simply hadn’t been considered. Today, with more and more connected devices every day, each and every one of them has to be secure to prevent an accidental entrance to the secure hub at the centre. Default device passwords are intended as a temporary measure and yet many of us never change them, (I for one have certainly never updated the Bluetooth password on my car, I’d never remember how to connect my music again). With more and more connected devices this approach is simply no longer an option, and Jason explained that this is why Arm is taking a revolutionary move to rethink security from the ground up and ensure it’s built into the system from the start. Secure cores and technologies like TrustZone have made an awesome start, and now we’ve also launched PSA, Platform Security Architecture, which is allowing us to release open source reference software and enable intelligent security, but we’ll hear more about that soon. No security will ever be 100% fool proof of course, and so we still need contingency plans. We need to know the very second there might be a breach so we can address it before any damage is done. Seemingly simple things like utilising a phone’s sensors to establish that it’s locked and upside down in a pocket, and therefore probably shouldn’t authorise that automatic pay request, are just the beginning. These newer, smarter chips effectively run a cyber healthcare system, constantly checking and monitoring the security of our systems and devices.  

When I spoke with Jason afterwards he was very impressed with his first visit to the Arm Tech Symposia: ‘Having just come from TechCon, it’s amazing to see another event on such a big scale and it says great things about the size and importance of our China ecosystem of partners.’

The newest Arm IP

Next up on stage is the Client Line of Business’s very own Ian Smythe to share with us the latest brand-new Arm powered products. Ian reiterated what Jason had shown, that with the Softbank journey we’re even stronger than we were 12 months ago and have reached the deployment of 100 billion units of Arm architecture only due to you and the amazing products you create. That first hundred billion has happened over 25 years, so when you consider that the next 100 billion will come in just four years it really is incredible. Arm is in the immensely privileged position of enjoying such strong partnerships which allow us to look ahead and see what is needed to solve the challenges the industry faces in the future of mobile, infrastructure and the embedded and IoT markets.

Ian Smythe looks at the future Technologies of Arm

We’ve always been confident about the evolution of the processor market but now we have to know how our partners will deploy across graphics, server networks, and that connected, embedded IoT which is set for exponential growth. We need to know how can we take the mobile journey we know so well and push that same consumer experience to all markets. To do this we need greater compute at the edge to deliver distributed intelligence. Mixed Reality is a key driver here, the convergence of the real and the unreal into one system is driving innovation around how consumers interact with their devices. As Jason said, we’re going to see a lot more intelligence from machines themselves, and this AI simply must be safe and efficient as we scale it from edge to cloud. AI is the broadest version of the concept, with Machine Learning being the area we typically focus on, giving networks vast chunks of data and expecting them to learn and make intelligent decisions. This isn’t far in the future though, Ian reminded us that we already use AI every day, with his kids talking to Siri or Google assistant, and expecting a response. To us, AI makes our lives simpler. I don’t speak or read Chinese, but Google Translate, optimised by Arm to work offline, has allowed me to consume a much greater range of content than would otherwise be available to me. This type of instant consumption ML needs to come off the server or cloud and be performed at the edge, and there are many reasons for this. For one, the bandwidth needed to keep going back and forth to the cloud is unsustainable, Google figured out that if everyone started to use voice recognition for just 3 minutes a day, they’d need to double their server farms. Cost and latency are also huge factors, autonomous vehicles can’t tolerate the latency inherent in a conversation with the cloud, decisions need to be taken instantly. So, how do we enable this at the edge? Ian explained that firstly, it’s about compute. DynamIQ is an Arm technology which supports high speed, low latency connection to AI accelerators. Introduced earlier this year, this is changing the way we work with added flexibility such as we’ve never seen. With the first DynamIQ based processors, Cortex-A75 and Cortex-A55, launched in March, we can support up to eight cores in a single cluster. Best of all though, they can all be different, handling different compute requirements at different power points for different use cases. One of the key requirements in building billions of units is matching performance to requirements, and this DynamIQ flexibility allows us to power everything from highly efficient, low power wearables, to the highest performance laptop type device.

As we know, it’s not just about compute. The Mali-G72, released at the same time as the latest Cortex cores, was also designed to support the highest performance use cases like High Fidelity Mobile Gaming, VR and ML on device. So, we’re already building next-gen premium mobile systems, but it still doesn’t stop there. Ian was pleased to announce the next step on this journey: an intelligent, complete Display Solution launched last week in Taipei. We’ve touched briefly on the complex use cases requiring these advanced systems, but for a bit more background, you can read about the disruptive technologies that made this so essential.

A complete solution 

The complete Arm Display Solution is made up of the Mali-D71 Display Processor, Assertive Display 5, and CoreLink MMU-600. Mali-D71 is the latest Mali display processor and designed to take the handling of complex use cases off the hands of the GPU, supporting ever higher performance across multiple use cases. Read all the details of the Mali-D71, formerly known as Cetus, here. CoreLInk MMU-600 goes with the Mali-D71 to add not only performance, but also a super secure connection protecting the content we consume. Read how MMU-600 achieves this here. In turn, Assertive Display 5 adds functionality specifically designed to deliver the most beautiful HDR experience across all panel types, even SDR ones. Here’s how.

Ian Symthe presents an Arm based system diagram

This complete Display Solution makes up the final piece of our latest Arm Solution. This premium system contains all the elements you need to deploy across a whole host of devices, and when you add in the safety island, it scales across even the most safety critical use cases:

So. Security. A vision of ours you’ll have heard before is to deliver a trillion devices. These wouldn’t be sustainable if they weren’t secure. This security has to run across every single link in the chain and we have to be able to trust it implicitly. This is critical to the ability to scale to the sheer size of the IoT. Luckily, we’ve been building security ecosystems for the best part of two decades, since SecureCore was first built for smart card payments. After that came TrustZone, which became a standard across the entire range of Cortex A processors. From this has grown an ecosystem of trusted execution environments that can run on Cortex, whether it’s mobile payment or streaming of high value, secure content. Our latest security solution is Platfrom System Architecture (PSA). PSA is a common framework for scaling connected device security. We’ve also added Cryptoisland-300, the first in a new line of security enclaves. PSA is our vision for how to secure and scale for the IoT and to do this meant analysing the treats and how to address them. Millions of developers across the world are working on the IoT, and this security is essential to the success of the ecosystem.

So, we’ve looked at not only the bright future of the industry, and specifically the incredible support of the Chinese ecosystem but we’ve also had a first hand view of the very latest in Arm technology bringing the future ever closer. Right now I’m off to hit the Tech Talks for a whole host more detail so check in tomorrow for the details!

Anonymous
Graphics & Multimedia blog