This is a liveblog of the Intel Developer Forum (IDF) Day 1 keynote. I was lucky enough to be invited to attend, as I did last year, and I’ll be liveblogging as many sessions and events as possible.
The keynote starts promptly at 9AM with Ulmont Smith (I didn’t catch his title at Intel). He mentions that this is the 16th year of hosting IDF, and previews some of the things that will be available this year at IDF: 170 technical sessions, poster chats, longer hours in the Technology Showcase, and more engineers on-site than any previous IDF. Ulmont also previews the attendee appreciation party (with Counting Crows) and gives a word of thanks to the IDF sponsors. Ulmont then introduces the new Intel CEO, Brian Krzanich, and the new Intel President, Renée James, who will be the keynote speakers in the day 1 keynote.
Brian Krzanich, the CEO of Intel, now takes the stage. He starts out with discussing what IDF means to him (and what it means to Intel). Krzanich talks about how this is an exciting time in the industry, and he indicates that he’ll lay out Intel’s strategy for how they will succeed in this highly transformative time in the IT industry. The underlying themes driving this transformation are related to the “connectedness” to the user; as computing moves closer to the user, the volume increases. So, in the transition from servers to desktops to notebooks to tablets to phones and next to the “Internet of Things,” the volume of computing increases. Computing is getting more personal and more connected, according to Krzanich.
Krzanich talks about how this migration to the “Internet of Things” drives Intel away from CPU-centric architectures into integrated (i.e., system-centric or System-on-Chip [SoC]) architectures, and he believes this is why Intel will win and drive marketshare in this transition. He talks about Intel’s assets—46K engineers, $10B in R&D, etc.—and how that will help Intel be successful. Intel’s plan is to lead in every segment of computing. That includes servers, desktops, notebooks, tablets, phones, and emerging segments.
With that, Krzanich takes a deeper look at some of these segments. He starts with the data center, and briefly discusses Intel’s leadership in CPUs (from the low-end with Atom and Avoton to the high-end with Xeon E5), server rack-scale architecture, and software-defined networking. He next transitions into a brief discussion of the evolution of the PC. He uses an HP prototype to talk about how notebooks will be fanless, lightweight, with long battery lives and ample computing power.
Not surprisingly, Krzanich next talks about Intel silicon. He announces a PC built on a 14nm SoC-based notebook, which Intel intends to start shipping by the end of the year (the 14nm SoC, not the actual products). “14nm is coming to a PC near you!” he says. Next he focuses on what he calls the “2-in–1″; these are the notebook/tablet convertible devices that easily switch between form factors. Krzanich believes that “2-in–1″ believes this is where the PC is headed.
But what about tablets? He picks up a Lenovo-branded Intel-based tablet to show that Intel-based tablets are available today. He points out that users can choose between Windows or Android, but of course there is no mention of iOS-based devices. Intel is targeting a sub-$100 system price point for the 2013 holiday season. This naturally leads into a discussion of what Intel’s doing in the phone market, and he shows off the first 22nm SoC-based phone, and discusses extensive LTE support—both for data and voice over LTE. Krzanich also shows LTE Advanced and carrier aggregation allows speeds up to 70Mbps (with 150Mbps quite possible).
Krzanich next moves to the mythical “Internet of Things.” Data volume, battery life, and security are all very important. This leads him to an announcement of the Intel Quark family of silicon chipsets. The Quark SoC is Intel’s smallest SoC, 1/5 the size of Atom and drawing only 1/10 the power. He next shows off “wearable” products that Intel is designing (based around Quark, I would assume). More reference designs around Quark are forthcoming.
Summarizing what he’s shown the attendees, Krzanich refers the Intel’s “landscape of opportunity” with products ranging from low-end server CPUs to high-end server CPUs, SoCs for tablets, SoCs for phones, and the all-new Quark SoC for the “Internet of Things” and wearable computing. With that, he wrapped up his portion of the keynote, and reminds attendees that he and Renée will be doing open Q&A at the end of the keynote. He turns the stage over to Renée James, Intel’s President.
James now takes the stage, and reminds attendees of Intel’s 45 years of leadership and innovation. She believes that Intel will help society transform to what she calls “integrated computing,” where we’ll move away from worrying about the form factor of computing toward using technology to transform lives and solve big problems. Her intent during this portion of the keynote is to show some projects that are already underway on how integrated computing will change computing. James gives us a quick review of some Intel history, reminding attendees that “Moore’s Law” remains alive and well. Krzanich announced 14nm today, but James points ahead to 10nm (in 2015) and 7nm (in 2017). James points to a number of technical hurdles that Intel has overcome:
- 3-D transistors
- Gate last approach
- Phase shift masks
Overcoming these technical barriers has allowed Intel to continue its leadership in computing. James shows off a cell phone manufactured using 1500nm technology. She compares that to a modern state-of-the-art phone. This phone, a Lenovo K900 (I think), runs at 2GHz and has more computing power than a Pentium 4 processor.
James reviews three “phases” of computing: task-based computing, lifestyle computing, and integrated computing. James’ discussion of integrated computing is inclusive of terms like “Internet of Things” and wearable computing; it’s a vision of embedding silicon/intelligence/sensors into everyday devices. She uses an example of how Dublin, Ireland, is using integrated computing (sensors in the city drainage systems) to intelligently manage drain water, traffic, and congestion in real-time. This leads James into a discussion of how integrated computing can be used to manage “mega-cities”.
The next use case of integrated computing is in healthcare. She shows off a wristband-based device that gathers health metrics in real-time. Next, James shows off a silicon-based patch that will replace the wristband device she just showed, and this silicon-based patch will gather health metrics in real time.
Of course, all of these devices are transmitting data, and this leads James to a discussion of big data. She uses human genome mapping as a example, and she talks about how the cost and time requirements for human genome mapping have dropped dramatically. Intel’s advancements in computing have driven the cost and time requirements for human genome mapping down to weeks (targeting days and hours) and down to about $5,000 (and targeting less than $1,000). Why is this important? Because at these cost and time requirements, it enables customized healthcare—for example, being able to create treatments that are targeted at an individual’s specific ailments (they used cancer as an example).
James closed out her portion of the keynote with a quote from one of Intel’s founders, Robert Noyce: “Don’t be encumbered by history, go off and do something wonderful.” At this point, Krzanich returns to the stage, and he and James open up for general questions and answers (which I do agree is unusual for a keynote).