 So we were here at the Computex 2018 and a couple days ago you announced A76 and G76 and V76, so now everything is 76. This is true, we actually have a premium IP suite that is fully aligned. As I mentioned I think in the launch and you've seen some of the information, we're delivering laptop class performance with the Cortex A76, really developing high fidelity gaming or actually cross platform gaming level support with the Mali G76 and the first 8K capable video processor in the V76. They put it all together along with the other key components that truly provides a much bigger platform for the next generation. If you merge in what we're doing with the Project Trillium for acceleration for machine learning, all of the key components do better machine learning on their own independently but also as a combination of all of the above. So I'm really happy that you're calling it a laptop class performance because I've been using ARM laptops for the last 5 years right, but they've been mostly Chromebooks and is it true to say that there hasn't been such a huge push, like you've been always talking about it right, but now are you pushing it even more? So it's a consistent move as you probably see from some of the collateral that we've generated. The trajectory that we have on performance has been going very strongly up and to the right but what the A76 represents is really a much bigger step as you point out. Over 35% over the last generation of product like the A75 but also nearly two times when you consider the actual delivered performance versus today's maybe A73 class products that you see. So two times faster than the ARM laptops that are there right now that are being talked about and shown off and stuff. That's a big jump. You're talking about it's like Intel Core i5, is it? So we believe that especially for the mobile connected laptop the sweet spot for us is versus the Core i3 i5 class devices and certainly the A76 has the capability to compete quite well with them while being substantially more power efficient. Because ARM processors have always been designed mobile first a little bit right now. Is it going to be a little bit more laptop first? Is this single thread performance? Are you focusing more on that or not necessarily? So we are talking about single thread performance. It will still be big little performance overall but we are looking at this device still to go into premium smartphones. So effectively while we will have that kind of performance available it is still looking at a smartphone form factor when it comes to power efficiency. Some of the most awesome smartphones that I see have this PC mode where you connect them to an external display and it has this PC UI and stuff. Hopefully this becomes more and more of a push. So how much is ARM doing into that field? So our focus has been enabling the level of performance and working with the ecosystem and make developers more capable of optimizing to ARM platforms. The final delivery of those types of devices naturally comes from our customers or partners and their customers. But we are actively trying to help in every way we can to make the ecosystem much more capable of delivering optimized solutions for these devices. Maybe through Leonardo and all this open source Linux optimizations for the ARM because there was this target to get 25% of all servers in 2020. Are you going to reach this? So quite a big shift Mr. I'll talk through it anyway. Certainly Leonardo is a platform for us to deliver better open source solutions that can go into mobile markets, consumer markets and infrastructure markets. But we're also working with gaming developers. We're working on actual optimized libraries such as the ARM compute libraries. We're looking at ARM's NN framework or a platform that fits into frameworks like Android NN and Cafe. So we're doing a lot to actually enable outside of Leonardo as well. The second question you had was about servers and how ARM could get to 25%. Now that is something that ARM is still working with our ecosystem to enable and we see a lot of traction potentially on the cloud side. Because I was mentioning the server and that target of 25. Is there any target that you might have for the laptop market? So we do want to get to about 90% of all mobile devices which would mean that we want to get a larger percentage of clamshell devices. In this case we'd say it'd be north of 20% of clamshell devices by 2021-2022. That's cool. How about the A76? How is it a different architecture compared to the A75? Is it being thought of in a different way? It is. It is a ground up new micro architecture and the approach that ARM has taken is to build micro architectures and step them through and have a parallel team go to maybe two generations out and build the next big step. So it does rethink a lot of what the micro architecture is. Naturally in micro architecture terms there's a lot of effort into newer branch prediction techniques, better prefetching techniques, greater and wider out of order issue to name a few things that are different about the Cortex A76 versus the prior generation A75. Is it already fixing all this security stuff that was happening with the high end of ARM and that's mostly Intel issues, right? That's also affecting high end ARM. Is it already fixed in A76 or they can't be fixed that fast? Just to be clear, the spectra and meltdown issues that you refer to, there are multiple variants and out of those variants two are especially variant one and variant four are primarily software issues that need software diligence to fix. So you can't address all of those just through hardware. For variant two and variant three Cortex A76 does not have any effect on variant three or does not, variant three does not affect it and we have our fixes in place for variant two as well. So by the time these devices come to market at least the large volume ones you should be kind of resistant to or have mitigations to variant two and three. And the fabs have been showing off that they have seven nanometers working so this is going to be a nice jump from 10 nanometers, right? So yes, there is a step up to seven nanometers. Remember that ARM still delivers synthesizable IPs so Cortex A76 would still fit into 16 nanometer or 10 nanometer type devices. But seven is a sweet spot especially when you consider premium smartphone and laptop or clamshell class devices and we're quite excited as a combination of these two, the process technology and the processor, what kinds of devices are possible. Because for example Qualcomm did a nice jump in their first and second generation 10 nanometers, is it going to be the same in seven nanometers or is it potentially going to be two different versions or something like that? So I can't comment on what Qualcomm does for sure so that's a question for them but we certainly see the broader ARM ecosystem being able to handle seven nanometers and we will see exciting devices as a result.