Last November Samsung said they were working on a dual core 2 Ghz Cortex A15 chip called Exynos 5250. A dual core 2 Ghz Cortex A15 chip should be at least 2.5x as powerful as the current dual core 1.2 Ghz Cortex A9 chip in Galaxy S 2. And that’s a difference that you will feel in real world performance, unlike with current quad core chips, where you might feel an improvement in performance, but not twice as big, and only for certain apps and games.
This chip will apparently use the much more powerful Mali T604 GPU, which should be about 5x as powerful as the current Mali, according to ARM. And it will apparently be used to power the resolution of the next-gen Galaxy Tab (running Android 4.0), which will features a breathtaking 2560×1600 resolution. This may sound unbelievable, but Samsung demoed this resolution on a 10″ panel last year in May, and said it would be available in the first half of this year. So all the pieces are falling into place.
AndroidAndMe is reporting that this tablet might be bigger than 10.1″. I’ve been wondering why Samsung and others are willing to take risks with smaller tablets than 10″, but they don’t want to risk making a bigger one. I can see how a 10″ tablet would be too small for people who don’t need to hold the tablet with one hand in air all the time.
For example, I think a 11.6″ tablet would make perfect sense for a Transformer-like hybrid device, because then it would be more like an actual ultra-portable laptop. Plus, I’m sure Samsung’s tablet will be even lighter this time around than their Galaxy Tab 10.1, which was 599 grams. I was very surprised by the weight of their Galaxy Tab 7.7, which is only 335 grams. That is crazy light, and they can probably make this tablet weigh only around 500 grams or so. This means, even if the tablet becomes a little larger, it wouldn’t be heavier.
Going back to the Cortex A15, Taylor is probably right about them being able to make it before TI’s OMAP 5 because it’s made at a 32 nm processing node, while the OMAP 5 is made at 28 nm, which wasn’t as ready and stable, but I think it’s also because TI is reaching straight for the maximum of 2.5 Ghz per core, while Samsung decided to leave it at 2 Ghz so they don’t have to work on optimizations for as long, and be able to release it early in the market.
That was a smart decision, because Cortex A15 will bring massive improvements over Cortex A9, and they’ll be the first ones to market it and sell it with their devices, while still being able to upgrade it to 2.5 Ghz by the end of the year, so it can compete with the OMAP 5.
The question remains, will this chip be in the Galaxy S 3, too? I wouldn’t be able to say either way right now, because on one hand Cortex A15 may be more power hungry, which is why they are putting it in a tablet perhaps. On the other hand Samsung has also said they will be able to ship Cortex A7 processors in big.Little configuration by the end of 2012.
So it’s possible the Galaxy S 3 will have a quad core processor made out of a dual core 2 Ghz Cortex A15 for very high-end tasks, and a dual core 1-1.5 ghz Cortex A7 that will handle most of the other tasks. But if they weren’t able to do it like that this early, perhaps it’s better they stick with Cortex A9 for the GS3. I probably wouldn’t want Cortex A15 in a phone until it’s ready to work with Cortex A7 as its companion core. But it’s all speculation for now, and it’s possible that a 32 nm Cortex A15 could be even more efficient than an overclocked Cortex A9.