Nvidia blew everyone away at last year's MWC when they announced a quad core ARM chip. By this time we almost take it for granted because we heard a lot about it and we know others are making quad core chips as well, including Samsung. To tell the truth, I kind of expected Nvidia to announce a quad core. I knew they are crazy enough to make a chip like that for smartphones and tablets. Nvidia's CEO is pretty great at marketing their chips and evangelizing them.
I was a big supported of Tegra 2, because it all sounded so good on paper, and Nvidia was going to be the first one to market – but mostly importantly they were going to be almost 6 months ahead of everyone on the market. That showed how aggressive Nvidia is treating this ARM market that they are already beating long time leaders such as Qualcomm and TI.
The Tegra 2 was supposed to launch sometime in early fall – and it didn't. It launched for the first time in February of the next year with the Motorola Xoom tablet and the Atrix (Optimus 2X was launched a month earlier in South Korea). If Nvidia actually launched it in September, it would've been a very big deal, because nobody else had a dual core in the market until around May or so the next year. But by the time the Tegra 2 was on the market, we already knew that the chips from competitors Samsung, OMAP and Qualcomm are going to be more powerful.
Unfortunately, for whatever reason, Nvidia repeated the same mistake (or technical difficulty if you will) this year, too, with Tegra 3. It was once again supposed to launch in August or September. When did we see it for the very first time in a product? At the very end of December in the Asus Transformer Prime. That was the time when were were supposed to see the phone version of the chip, because Nvidia makes the phone versions about 3 months later compared to the tablet versions, unlike other manufacturers. This is why some of us were hoping we'd see Tegra 3 in the next Nexus phone.
To make the story short, we're once again in the same situation where by the time Tegra 3 gets to be in several products, there are already other chips that promise to be better – quite a lot better in some cases. How is that possible? Well for once, some manufacturers chose to continue the dual core path, and get the most out of it, instead of going 4 cores.
This will potentially hurt them marketing wise, because "quad core vs dual core" sounds a whole lot better than "more powerful dual core vs dual core", but they should win the performance race in most cases, because the new dual core chips will be more powerful in those 2 cores, while the Tegra 3 will be weaker. Since most apps are still not using all 4 cores, this means that for now dual core chips will still be more important than 4 cores.