Intel Debuts Clover Trail For Tablets, Launches New Atom Inside 88
An anonymous reader writes "Today, Intel is launching its next-generation Clover Trail platform. The new Intel Z2760 is a dual-core, quad-threaded device clocked at up to 1.8GHz, with support for up to 2GB of RAM and graphics provided courtesy of a single PowerVR SGX545 core. Chipzilla expects to see wide adoption from multiple partners, with a host of tablets expected to launch simultaneously with Windows 8. The new SoC is closely related to Medfield, Intel's 32nm smartphone platform that ExtremeTech reviewed earlier this year, but there are a few differences between the two."
Re:Semi-Accurate predicts horrible failure (Score:4, Insightful)
I don't know whether that article is correct or not; but(if we assume for the sake of argument that it is) the thing that really doesn't make sense to me is what Intel's motivation would be:
He gives the impression that both Intel and Microsoft are working against the part(Intel doesn't want the lousy margins, Microsoft is undermining Win8 x86 tablets with office licensing terms); but that the part is also a 100% hardware-locked MS-only device that they won't even let any other potential customer have a crack at, much less have any genuine enthusiasm for trying to sell it to them.
However, if that is so, why does the part exist at all? If it is so big that it is practically an i3, why would Intel design an entirely different chip rather than just lasering something off their weaker i3s and giving them an inscrutable model number and OEM only distribution? And if Microsoft is shafting Intel on the part, why is Intel making it a Windows exclusive(and on the processor level, not merely with a locked bootloader, which would allow MS OEMs to do their thing and Intel to still sell this to Android vendors, embedded linux appliance vendors, and whoever else)?
That's what I don't understand about the story as presented: plenty of products ship deliberately crippled because somebody prefers them that way(see all DRM systems, for instance); but that crippling isn't free, so it has to be in somebody's apparent interest for it to happen. Here, none of the players seem interested: Intel, allegedly, has a chip that can't compete with their existing products, and isn't a winner on margins. MS has a chip apparently designed just for them, except they would rather ship punchier devices with those existing products. So, who wins here?
I think people get confused by the "semi" part (Score:4, Insightful)
His blog would better be known as "Rarely Accurate". Every once and awhile he gets good information but usually, he just makes shit up.
For example he claimed that Kepler, nVidia's architecture now known as the GTX 600 series, would be under-performing and feature large parts dedicated to PhysX and that AMD's CGN would be a "clear winner". As it turns out that is not the case, the GTX 680 is extremely fast, quite energy efficient, and beats out the 7970 from AMD. And no big PhysX areas, just lots of stream processing cores (which is what GPUs are these days).
Then when it launched and had big shortages due to extremely high demand he claimed it was because it didn't work right on TSMC's 28nm process and supply issues would forever plague it. Now, they are in stock everywhere and can be had in quantity whenever you wish.
I really wish people would stop linking to this troll and giving him ad money. The guy is not a great source of information and worse still in the absence of actual information he'll just make shit up.
The thing certainly doesn't use a ton of power, it's TDP spec is less than 2 watts and I have never seen Intel underspec a TDP. That means it will never, under any situation, dissipate more than 2 watts and by extension never draw more than that.
The question is, of course, how well it performs at that power level, particularly compared to competition at the same level.