Intel revealed today at IDF that its 32nm processors, previously known by the codename Sandy Bridge, will be called “2nd Generation Intel Core” but will still go by the product names of i3, i5 and i7.
The new processors should appear in both desktops and laptops early in the new year, so we’ll expect to see a swag of them announced during or just after the Consumer Electronics show in early January 2011.
So what’s new and exciting?
There’s a new “ring” technology to integrate graphics into the processor die, and it will allow the sharing of fast cache between the cores of the processor and the graphics processor unit – which handles 3D and media processing .
As an evolution from SSE graphics instruction sets, Intel now has AVX (Advanced Vector Extensions), that perform better with floating point operations – meaning that digital photo editing and similar tasks should get faster.
In the demo we saw at the keynote, it seemed that for advanced photo processing and video rendering a 2-5x speed up should be expected.
A games demo highlighting some of the power of the new integrated graphics showed off Starcraft running on integrated versus discrete graphics.
The new "Ring" architecture to integrate graphics and CPU means that both can use Turbo Boost and share memory and cache.
While the results were impressive, it was noticeable that as the workload on the graphics increased, the discrete processor pulled ahead in the framerates substantially. We’ll look forward to testing the new processors on our own benchmarks to see just how good the new integrated graphics really are.
TDP? That’s just an estimate, now
One of the key features of the new Sandy Bridge processors from Intel will be the extension of Turbo Boost technology.
While in the current generation Westmere processors (Core i5 and Core i7) turbo works exceptionally well when a single-core task is demanding more power than normal, it didn’t work so well for multicore workloads. So the more cores you had, the less Turbo Boost benefits you got.
The new Turbo Boost is able to exceed the TDP limit, which means that multicore workloads now get the benefits. Intel says that a new power-averaging algorithm manages power and TDP to optimize performance.
(Of course, we are left wondering what TDP means now, if exceeding it is standard.)
In addition, now that the ring structure for the graphics and processor cores is in place, the CPU cores and graphics subsystem can both use Turbo Boost – though only one subsystem can use it at a time.
If you need the compute power from graphics – such as when rendering video – the graphics subsystem will use Turbo Boost to improve throughput, while if processing is required, the CPU will get the benefits.
We’ll be interested to see how this works in real-world situations.
At Intel's developer forum in San Francisco, the enthusiasm of the attendees is matched only by the jauntiness of the slogans plastered on the posters.
Zara Baxter travelled to San Francisco courtesy of Intel.