Intel is expected to shed light this week on its next generation of chips.
The chipmaker's premier event of the year, the Intel Developer Forum (IDF), is expected to draw several thousand hardware developers, Intel partners and media types to San Francisco from Tuesday to Thursday. Intel usually holds two IDFs in San Francisco annually, but moved the spring event to Beijing this year to highlight the company's growing China operation and cut the costs of holding two major US events.
Chief executive Paul Otellini will get things going on Tuesday morning, and he's expected to preview Intel's upcoming Penryn generation of processors, which will use a new, smaller 45nm (nanometre) manufacturing technology (due later this year). Later in the week, other Intel executives are expected to highlight the company's growing interest in low-power mobile and consumer electronics devices that could help Intel cope with a rapidly maturing PC market.
Intel is on a roll this year, following a dismal period in which the company lost significant market share and design credibility to AMD. AMD's Opteron chip was the processor of choice in the server market right up until the middle of last year, when Intel introduced dual-core chips based on a new design and then combined those chips in a package to make quad-core processors. Those quad-core chips, combined with an unrelenting price war, have had AMD on the defensive for much of 2007.
Barcelona, AMD's quad-core processor that launched last week, will help the smaller chipmaker regain some momentum, but it's not the knockout blow AMD wanted to land when it started developing Barcelona. Intel hopes to regain a decisive lead with the Penryn processors, which are expected to arrive by November.
So, expect to hear a lot of information about the Penryn chips, such as specifications, power consumption and benchmark projections. Intel will also likely tout its manufacturing prowess in getting to the 45nm level well before AMD is expected to do so in mid 2008. The smaller manufacturing technology makes it possible to cram more transistors onto each chip and also allows Intel to cut more chips from each 300mm silicon wafer it produces, which lowers the overall cost for making each chip and makes the accountants happy.
New directions
But while Intel is naturally very concerned with shoring up its position in its core business, it's also trying find the next big source of growth for the company, which is taking it in some directions outside of its historical expertise.
Intel has been desperately trying this entire decade to find a sustainable business outside of the PC market. PC shipments are still growing at a relatively steady pace, especially in markets outside the US and Western Europe, but they aren't growing at a pace that makes Wall Street giddy. Intel spent much of the early part of the decade talking about the pending convergence of communications and computing, but the company isn't really participating in that convergence outside of its core PC business.
Intel's main success on the communications front has been Wi-Fi. The company convinced millions of people that its Centrino brand means wireless technology, even though it really stands for a package of chips, including the processor and a Wi-Fi chip. But Intel doesn't sell any Wi-Fi chips that aren't tied to its Core 2 Duo mobile processors — meaning its communications efforts are completely dependent on the health of the PC market. Otherwise, Intel has little to show for its turn-of-the-century spending spree on several communications chip companies.
So, it's now looking for the next big thing in two places: low-power chips for mobile internet devices (also known as UMPCs) and WiMax, a wireless technology with a much greater range than Wi-Fi.
In 2008 Intel plans to introduce a new x86 processor, code-named "Silverthorne", that the company has slated for mobile internet devices (MIDs). Intel thinks Silverthorne will allow MID developers to come up with more compelling designs that could help bridge the gap between smartphones and PCs.
The title of the "One True Mobile Device" is still very much up for grabs, and Intel is highly interested in making sure it's inside whatever device winds up winning that contest. At one point it hoped to participate in the mobile-phone market with a combination of a low-power applications processor and a cellular chip, but it jettisoned its mobile-phone division last year after its cellular chips were released in a grand total of two commercially available mobile phones.
Now, Intel's focused on this idea of a mobile minicomputer that's more powerful than a mobile phone, but more portable than a notebook. Several different MID designs should be on display at IDF this week, and Intel will also probably have concept designs based on Silverthorne to show off to attendees.
Executives will also probably spend a fair amount of time promoting Intel's work on the evolving WiMax standard. Intel sees WiMax as a way of bringing high-speed internet connections to homes without having to rely on physical cables, but it's an uphill battle against the entrenched cellular industry. Sprint has agreed to deploy WiMax in the US, but it's far from clear whether that carrier has the ability to deliver on that promise. Intel is also backing a start-up called Clearwire that's building a WiMax network.
Read this
Leader: Four core years... and then?
AMD's Barcelona keeps the company in the multicore race. That may not be the right game to play…
Interestingly, there won't be a presentation by Eric Kim, head of Intel's Digital Home Group. An Intel representative said there will be announcements related to Intel's home technology products sprinkled throughout other speeches, as the company also tries to get its chips inside a new generation of set-top boxes and media-friendly PCs. But, with its decision to put more weight behind its Core processor brands, Intel seems to be de-emphasising its Viiv home-entertainment PC brand.
With Intel's clear roadmap for the future and a relatively problem-free year, it has plenty of reasons to be confident, said Dean McCarron, principal analyst with Mercury Research. "They've come a long way since they had their issues two or three years ago," McCarron said.
