Intel today is hosting the AI Everywhere event, and we're liveblogging it! The company continues to be a technology leader, with the largest market-share in PC and data-centers alike. The company sees great potential in AI acceleration on the edge, with AI making its way to mainstream consumer applications. It knows that not all AI workload can be processed in a datacenter, and is empowering its next generation Core Ultra "Meteor Lake" mobile processors with a powerful neural processing unit, the next evolution in its Hybrid CPU architecture, and a feature-rich new iGPU that can take the fight to AMD's expertise in the field. Later in the event, Intel is announcing its next generation Xeon "Emerald Rapids" server processors that dial up CPU core counts per socket, use the latest CPU core architecture, and are packed to the gills with server application-specific fixed-function accelerators. Join us as we unwrap this evening's biggest announcements.15:00 UTC:Intel is streaming from NYC. "AI is everywhere.

15:03 UTC: Gelsinger is on a rooftop, next to an active construction site.15:03 UTC: Welcome to the silicon, about 25% of our economy is built on silicon; which will grow to 33% thanks to AI.15:05 UTC: AI is "augmenting intelligence."15:06 UTC: "It's our job to make a neutral technology a force for good."15:07 UTC: Intel is accelerating AI everywhere-—in the datacenter, the edge, and PC.

15:08 UTC: "We are scale, we build AI into every platform"15:09 UTC: Intel advocates AI on the edge, on the device.15:10 UTC: 2024 will mark the AI PC.15:11 UTC: Intel's AI push is akin to Centrino—which mainstreamed wireless internet on the PC

15:12 UTC: Intel's foundry ramp up on track. 5 nodes in 4 years15:12 UTC: Dell, HP, Lenovo, Super Micro ,and Microsoft give their vote of confidence for Core Ultra and the AI PC of 2024.15:16 UTC: AI PCs will make up 80% of the PC market-share by 2028, according to BCG.15:18 UTC: 100 million client processors over the next 2 years: Intel plans to ship.

15:18 UTC: "CPU cores, iGPU, and efficiency matter"

15:19 UTC: Here it is, the Core Ultra "Meteor Lake"15:20 UTC: A 40% energy efficiency gain generationally, thanks to its next-gen hybrid architecture.

15:20 UTC: NPU is 2.5x better than previous-gen AI acceleration attempts by Intel (GFNI)

15:21 UTC: Core Ultra breats Ryzen by "double digits" (read: 11%)15:22 UTC: 79% lower power than AMD Ryzen 7040 series15:23 UTC: Intel Arc graphics beats AMD's Radeon 780M

15:24 UTC: The iGPU offers a 200% perf/Watt uplift over "Raptor Lake" iGPU15:25 UTC: The NPU is the star attraction, for power efficient AI inference and DNN building.15:25 UTC: Core Ultra is designed for local instances of popular generative AI.15:26 UTC: Intel NPUs are the most feature-rich and compatible across AI workloads.15:27 UTC: Intel finally has a solid ISV ecosystem for client AI15:28 UTC: AI will scale the fastest because Intel will deliver it.15:29 UTC: You can buy Intel Core Ultra today!

15:29 UTC: Audacity (audio app) gets generative AI support, to add instruments and tunes.15:31 UTC: Over 230 design wins.15:33 UTC: The largest architecture change in 40 years:15:34 UTC: The actiion shifts to data-center/server business.

15:35 UTC: 4th Gen Xeon Scalable with AMX was well recieved by the industry, Intel says.

15:36 UTC: Behold, the 5th Gen Xeon Scalable "Emerald Rapids"15:36 UTC: Xeon is the best data-center CPU for AI: Intel. 64-core/128-thread, with a dozen accelerators and AI-relevant extensions.15:38 UTC: IBM uses Xeon to accelerate WatsonX data store. 2.7x throughput upgrade for WatsonX.data15:39 UTC: Up to 77% TCO improvement when upgrading from 3rd Gen or older Xeon Scalable systems.15:40 UTC: Google Cloud has implemented 5th Gen Xeon15:41 UTC: Proxi is using Xeon Scalable to accelerate its LLMs without GPUs15:45 UTC: Intel announces Gaudi 2 generative AI accelerator. 50% perf/Watt over NVIDIA H100.15:47 UTC: Cloud-to-Edge-to-PC15:48 UTC: "Emerald Rapids is simply the best CPU for AI"16:00 UTC: Intel Gaudi 3 unveiled, coming 202416:01 UTC: And that's a wrap!