 My name is Frédéric Lucam. I work with ST Microelectronics. I'm a product line marketing manager of the STM32 H7 which is recently announced a new STM32. So we are adding this device. It's a major addition to the STM32 portfolio actually. It's part of the high-performance series and this device comes with a Cortex-M7 on it running up to 400 MHz of frequency. Thanks to that, we are able to reach a 2010 core mark of CPU computation capability. And it is this technology that we are using, which is actually a 14 nanometer technology, is allowing us to bump the frequency up to that point, which is a record performance in terms of the Cortex-M based MCU. This device comes with plenty of features on chip. So we have a large memory. So we have a 2 megabyte of flash on board, which is split in two blocks of 1 megabyte. We have a large RAM, which is a 1 megabyte of RAM, which is a scattered RAM, which allows, for example, to make some data transfer from external memory to internal memory and to use very intelligently the master DMA that we have in this chip. So we have a master DMA, which is supporting new features like the link list, for example. We have also standard DMA in the product that are the typical DMA that we used to find on some STM32. And we have also a basic DMA, which is located in a specific part of the product in order to optimize the current consumption. We're going to see that in a minute. In addition to this, we have a certain number of peripherals that have been completely renewed. So for some of them, it's a completely new generation of peripherals. We have, for example, a flexible data rate can. We have two instances of this on the chip. And we have also many analog features. We have a fast 14-bit ADC on board. Three instances of this ADC can be found on the chip. And we have also a couple of analog comparators as well as operational amplifiers. Back to the power consumption. So thanks to this architecture that we have implemented inside the product, we are able to get many different low-power modes. So we'll speak about that in a minute. But before that, I want to mention that we have a capability in run, which is in the range of 280 microns per megahertz, running full speed. So that's an amazing dynamic value if you consider the performance of the device. We have also these large configurations possibilities in terms of low-power mode that allow us to reach in the range of 8 microns in a standby mode, for example, which is very useful for application where you need to shut down or to minimize the current consumption or to avoid to drain the battery or just optimize the current consumption that you need to do in many, many applications now. The last point I want to mention before we go to the demo is about the security features. So this device also comes with some new services that we have implemented in a chip that will be able to serve two kinds of use cases. One use case is to be able to program the device in production in a secure manner, so which means that it will make sure that the programming goes well, that the device has been authenticated before programming, and it means that you will put the right code inside the device. And the second use case that we are targeting with these new services that we have implemented in the product is the firmware update while the device is already in the field. In the same manner, the customer will be able to develop a scheme, I mean a software that will run in a secure area of the device in order to make sure that the reprogramming in the field can be done in a secure manner. So now I suggest that we go and see a couple of demos that are running on this evaluation board. So now let's look at a couple of demos that are running on this evaluation board. So that's a board that is roasting the STM32-H7 silicon in the superset version. So that's a 2 megabyte of flash and 1 megabyte of RAM. So that's the demo starts with this screen, which is where we can see that we have several types of demo that can be launched. It's just needed to push the button on it. And that's it. We see that we have the programming of the demo. So we'll see in a minute what the programming of this demo, for example, is finished and we can activate it. Okay, now what we are seeing is the CoreMark benchmark demo. So we have been running the CoreMark test suite here on the micro, running the core at 400 MHz. And you see that we get a result at 2010 CoreMark of computation power, which is an amazing value and set the record in terms of performance for an STM32 and for an MCU Cortex M-based micro. So the code is running for this demo. The code is running off of the flash and the data are stored in the DTCM RAM. The L1 cache is on. So we have actually two, we have a 16K byte of cache for instruction and 16K byte of cache for data. And the cache is on for this demo. Okay, now we are looking at a second demo, which is the MJPEG demo. So basically we are running four videos at the same time on the board. So as you can see, we have four different independent videos that are running and that's based on the MJPEG. So which means that basically we are getting the data from the external Quadspy, in particular the movie. They are decompressed by the MJPEG codec that is part of the MCU. And we are able to display those four videos using the ChromeArt Accelerator, which is on board the chip, which is able to speed up the send of the data to the display. Those four videos are running at 25 frames per second each. And as you can see here, when we look at the CPU load that we have running those four videos in parallel, we have only 1 to 2% of CPU load, which means that basically the CPU is doing nothing and all the tasks are executed by the hardware acceleration that is on board the chip. If we switch the large screen on one of these videos, we still see that we have very good rendering of the image, still very low CPU load in the range of 3% in this case, still at 25 frames per second. And we have also the possibility here on the demo to deactivate the MJPEG accelerator. So now I have turned off this MJPEG accelerator and we see that the rendering of the movie is not as good as it was before. The frame per second has been down to around 15 frames per second, as opposed to 25 in the other example. And the CPU load is now reaching in the range of 75, 76% just because we are doing the decompression by software. Now I have reactivated the MJPEG acceleration and we see that we are back to the previous results. I hope you enjoyed the demo. If you want to know more about the product, you can go on the ST.com webpage and you will get more data about the product. Thank you very much.