Chipzilla and ARM are going to war over the ultrabook market with both sides thinking they could win it.
The territory is close enough to a PC for Intel to think it belongs to them, and close enough to a tablet for ARM to have a chance.
Intel’s answer is to release new low voltage models within its Core processor family to target low power, cloud-oriented ultrabooks. What is interesting is that Intel has usually had a half-hearted stab at this market with Atom chips.
Ultrabooks are being seen as a more sensible approach to the keyboardless notebooks which are tablets. It is hoped that they can bring some new life into the fast fading netbook segment.
The yarn is that Ultrabooks can provide PC performance, using the Sandy Bridge processor rather than Atom, while keeping the mobility and long battery life, plus a width of under one inch.
Price wise they will hit the shops with tags of about $1,000.
To do it, Intel has had to make speed cuts to its new Sandy Bridge models. They run much slower but they have half the power consumption of the normal Sandy Bridge chips. They only need 17 watts compared to 35 watts.
The three new processors are all dual-core, with the same on-die graphics, and comprise the 1.7GHz Core i5-2557M, 1.7GHz Core i7-2637M and 1.8GHz Core i7-2677M.
Intel’s first hardware to use the technology is the 11.6-inch Asus UX21 which will be 0.7 inches thick and have a 64GB or 128GB solid state drive that can resume from sleep within a couple of seconds, which is better than we can do.
We will see this in September. Lenovo has been showing off the IdeaPad U300S and LG has the P220, both with 12.5-inch displays. Let us not mention the Apple MacBook Air, which will also be upgraded soon to the new Core processors. No, seriously, let us not mention it.
All this is a dry run for the release of Ivy Bridge which will be Intel’s first 22nm processors, followed in early 2013 by Haswell.
These will bring down the power rating to 15 watts, and mean that Intel will have its feet well and truly under the lower power table.