The Intel CES 2024 Pat Gelsinger Keynote Live Blog (Starts at 5pm PT/01:00 UTC)


This evening is the biggest PC-related keynote of CES 2024, Intel’s “prime” keynote with CEO Pat Gelsinger. Part of Intel’s “AI everywhere starts with Intel” campaign for the show, Gelsinger is expected to talk about the role AI will play in the future of consumer technology, along with the economic implications.


07:58PM EST – Thanks for joining us for our second keynote live blog of CES 2024

07:59PM EST – Pat Gelsinger’s keynote is scheduled to begin in just moments

08:00PM EST – Gavin is at the show, taking in the views from the keynote audience

08:00PM EST – While CES no longer has a true “prime” keynote as it did in the days of yore with Microsoft, it still has a few keynotes that are higher profile than others

08:01PM EST – And having the CEO of Intel speak is one of them. This is certainly the highest-profile of the PC-related keynotes at this year’s show

08:01PM EST – We’re not expecting any grand product announcements from Gelsinger in this hour-long keynote. All of Intel’s chip announcements were yesterday

08:02PM EST – But rather, this will be a higher level look at things about product strategy and the future of computing. And a whole lot of AI talk

08:02PM EST – And here we go, starting with a video

08:02PM EST – “CES 2024 is the celebration of all that tech has to offer humanity”

08:03PM EST – Opening the keyote is one of the CTA’s executives

08:05PM EST – Quickly recapping Gelsinger’s history for the general audience before having him walk on stage

08:05PM EST – He will be joined by a CNBC reporter, Kristina Partsinevelos

08:05PM EST – And here’s Gelsinger and Partsinevelos

08:06PM EST – Reiterating that Intel is in its financial quiet period right now. So the company and its offers have limited leeway to talk about things

08:06PM EST – Gelsinger classifies himself as an optimist

08:07PM EST – AI can be shaped as an incredible force of good

08:07PM EST – Comparing the utility of AI to not using spreadsheets in this day and age

08:07PM EST – There is a significant emphasis in this talk on “for good” versus “for bad”

08:08PM EST – Partsinevelos: “Can you trust a for-profit company to do good?”

08:08PM EST – Gelsinger: part of Intel’s brand is to be trustworthy

08:09PM EST – Open technology access is part of the key, so that people can see how AI works

08:09PM EST – “We love it when people copy us”

08:09PM EST – Pivoting to the subject of “AI PC”

08:10PM EST – What makes a server an AI server?

08:10PM EST – “Are you able to achieve a certain level of performance on AI workloads?”

08:11PM EST – And also the difference between AI training and inference. The former having a much stiffer requirement versus what can be done with inference in just a CPU

08:11PM EST – Though Intel is building products for both ends of the spectrum. Including accelerators for the high-end of the market

08:12PM EST – And there’s a recent trend to pare down models to achieve similar results without requiring upwards of a trillion parameters

08:12PM EST – Gelsinger doesn’t believe there’s an end in sight for the science of AI

08:13PM EST – “This rivals some of the early days of the PC; the early days of Moore’s Law”

08:13PM EST – Now on to consumer PCs

08:14PM EST – Should the average Joe care over the next 5 years?

08:15PM EST – Gelsinger is comparing the arc of AI to that of Wi-Fi and their Centrino initiative

08:15PM EST – Adding hardware capabilities now is going to bring about changes

08:15PM EST – Intel has a lot of competition in the AI space. What makes Intel different?

08:16PM EST – Intel will win by being first, by their volumes, and by their engagements with partners

08:16PM EST – “I believe this is a defining moment for the PC”

08:17PM EST – Being asked if there’s anything else to say about Gaudi 3?

08:17PM EST – Gaudi 3 is out of the fab now, for early test debug. OEMs and customers will be brought in to the debug process in the next few months

08:18PM EST – So far the chip looks “exquisitely good” ahead of its launch later this year

08:18PM EST – Gelsinger is applauding his Israeli teams for their performance, especially given the strife going on there

08:20PM EST – Gelsinger believes this is the last correction cycle for auto parts for the post-pandemic supply hiccups

08:20PM EST – Though it’s certainly still a bit of an uncertain economic environment

08:21PM EST – Discussing the economic status of various world regions, including China, Europe, and the US

08:22PM EST – “We’re more optimistic as we make it through the year”

08:22PM EST – Will LLM training catipult into a lot of lawsuits?

08:23PM EST – Gelsinger thinks a legal framework will get established

08:23PM EST – A trained LLM is basically compressing the Internet into a model

08:24PM EST – One way or another these LLMs will need content. Whether it means licensed or unlicensed content

08:24PM EST – AI is moving quickly. But politicians move slowly

08:25PM EST – “How do we come up with the regulations in such a short amount of time?”

08:25PM EST – Gelsinger thinks cases will eventually hit the courts, and that will start to bring answers and legal precident

08:27PM EST – The world is motivated to figure out AI by the productivity possibilities

08:27PM EST – The world will find value in AI, domain by domain

08:27PM EST – Over a 10 year+ cycle

08:28PM EST – And Gelsinger thinks that edge devices will play a huge role due to the privacy and cost benefits of not going to the cloud

08:28PM EST – Not to mention regulation once the cloud is involved

08:29PM EST – “How do companies like Intel make money” if this AI processing is all local?

08:30PM EST – Gelsinger: new use cases are going to bring new economic implications

08:31PM EST – And that’s a wrap on Gelsinger’s interview

08:31PM EST – Thanks for joining us!

 

We will be happy to hear your thoughts

Leave a reply

Funtechnow
Logo
Compare items
  • Total (0)
Compare
0
Shopping cart