 I'm very happy to be here today. I'm Kendall Nelson, like they said, from the Open Infrastructure Foundation. Joining me today, we have Alex Song, a contributor from Cyborg. Alex, maybe you can tell everybody a little bit about what Cyborg even is before we jump in. Yes, I will introduce Cyborg Projects. Cyborg is accelerator management projects with management, GPU, IPGA, SSD, and so on, as well as hardware, and software such as DPDK and CPDK. Cyborg Project has three micro-service, several APIs, several conductors, and several agents. Cyborg API provides API for resources, such as devices, deployables, device profiles, and so on. Cyborg Conductor connected to the device to, and Cyborg agent runs on computer node, which contains accelerator drivers, such as GPU, VGPU, driver, IPDA, driver, and the MSSD driver, and so on. We can set the drivers on computer first, and Cyborg agent periodically finds the devices and gets the device info. Then sends this info to Cyborg Conductor. Cyborg Conductor compares this info with the old devices, and updates these two devices, and this Cyborg can communicate with other components, such as NOVA, Neutron, Glance, and Sender to use these devices. Awesome. Thank you so much for that summary. I know Cyborg is a very bigger, up-and-coming project and certainly complicated in how much it has to work with NOVA and Neutron to focus on that hardware acceleration service. So I know that there's been a lot of work done between Cyborg and NOVA, particularly in the Xena release. Can you talk a little bit about the hardware enablement that as far as what happened there, this last release, and what also is coming up in the yoga release? Oh, yes. Cyborg products will have some interference and operation optimus and many drivers to support, such as silence, the IPDA, and internal pre-processed memory. And the enhanced about the accelerator, accelerator instance to migrate and resize. And we help team is the API filter, API filter conditions and the accelerator bound process, and so on. Thank you. Okay. A lot of work being done in the drivers and more focus on hardware acceleration, which is what Cyborg is all about. I know some of the work that went into the hardware enablement was around SmartNICS, and I hear that you have a demo to talk through for us today. Can you explain what is going on in that demo? Yes. Firstly, from the demo we can see that firstly, we need to create a device profile. The device profile is many kinds of accelerators combination, such as SmartNIC and NGPU. Then the NOAA API puts the instance with device profile. NOAA conductor calls NOAA scheduler to select available accelerator host. NOAA scheduler called placement to get the allocated candidates. Then NOAA conductor calls NOAA computer to put the instance. NOAA computer gets the instance and accelerator releases from Cyborg API and so it's this info to its XML file. Then NOAA, the instance can envisage visits as hardware from the host. Yeah, it's a lot of effort has gone into getting this to land between NOAA and Cyborg in particular, but also a little bit of work on the neutron side. So it's very cool to see all of this come together and as a part of the Xena release. So we're so pleased to be able to share what's been going on in Cyborg in both Xena and yoga and we really appreciate you all being here today.