The next, big thing for Intel: Sandy Bridge
Simply put, Sandy Bridge is an extension of Moore's Law that will take Intel chips to the next level of integration and performance.
Sandy Bridge is arguably Intel's most important future technology. So, what is it exactly?
Intel has been careful to reveal only snippets about the technology over the last 12 months or so. But enough is out there now to understand how the technology moves Intel forward.
In a nutshell, Sandy Bridge is Intel's next microarchitecture, or redesign, of its processors. A chip revamp is the single biggest undertaking for Intel. And it happens every two years. The current design, Nehalem, was introduced in November of 2008 and it pervades all Core i3, i5, and i7 processors (the latter two finally made it into Apple laptops on Tuesday). Its successor, Sandy Bridge, is scheduled to go into production in the fourth quarter.
While Intel Executive Vice President David Perlmutter said he would "not do a deep dive" on Sandy Bridge in his Intel Developer Forum Beijing keynote this week, he did reveal some key points about the architecture.
- More efficient: the central processing unit, or CPU, delivers a "significant improvement in instructions per clock," according to Perlmutter, meaning that it is more efficient at executing tasks.
- Faster on-chip communication: different parts of the chip will talk to each other faster--what Perlmutter called "improved inter-buses."
- Shared memory: on-chip memory called cache is shared between the CPU and graphics processing unit, or GPU.
- GPU now part of CPU: Intel combines the CPU and GPU on the same piece of silicon. According to an unofficial photo of the Sandy Bridge chip from Japanese Web site PC Watch, the GPU takes up roughly 25 percent of the processor's real estate.
- New instructions: Sandy Bridge will be the first chip to support Intel's Advanced Vector Extension (Intel AVX) instructions. AVX accelerates a host of multimedia tasks, including video and audio processing.
- More intelligent overclocking: and, finally, Perlmutter mentioned improved Turbo Boost--which speeds up (i.e., "overclocks") or slows down individual cores to meet processing or power efficiency needs.
Intel isn't being coy about where Sandy Bridge will play a big role: laptops--which will be more power-efficient while crunching video and audio data faster than the newest Core i7-based laptops today.
To drive this point home, in Beijing Perlmutter showed a Sandy Bridge-based laptop doing highly complex medical imaging, an application that would typically be done on a desktop PC. Analysts say this will set Sandy Bridge apart. "Intel mobile stuff used to be just a dressed down desktop processor. But that's changed," said Ashok Kumar, an analyst at investment bank Collins Stewart, who said the chipmaker is now challenged to strike a delicate balance between traditional, raw processor performance and how many additional features it can cram onto one piece of silicon.
And one of those additional features is a graphics processing unit, or GPU. For a mainstream PC processor, Sandy Bridge will be the first time that Intel puts a GPU on the same piece of silicon as the CPU. In other words, Intel's definition of a CPU now subsumes the GPU. Not a minor detail in the eyes of competitors such as Nvidia and Advanced Micro Devices, which together supply virtually all of the standalone "discrete," i.e., non-Intel, GPUs used in PCs today.
Nathan Brookwood, principal analyst at Insight64, points out that both the CPU and GPU in Sandy Bridge will be built on Intel's latest 32-nanometer technology, also a first. "This should help graphics performance," Brookwood said, adding that today Intel GPUs lag the CPUs by a generation. Both the CPU and GPU will share on-chip memory resources, too, also a first.
Put more simply, Sandy Bridge is an extension of Moore's Law--packing more transistors onto a chip. These days Intel is calling this "integrate." Take what was once off-chip--such as the GPU--and put it on-chip.
The goal of integration is to not just to build the fastest chip, but to build a more balanced chip. "The question is where do you compromise with the real estate that's available, as opposed to just making the fastest solution," said Kumar.