3 ms·
I'm somewhat excited about the "Convertible" concept. But technically, I don't think convertibles are quite ready yet for mainstream adoption. Frankly, the rea
by dragontamer 14y ago
I'm somewhat excited about the "Convertible" concept. But technically, I don't think convertibles are quite ready yet for mainstream adoption.
Frankly, the real problem is that Intel / AMD chips are not ready to compete in this form factor. Intel is making bold new moves with Haswell however, and AMD is making big moves with Kabini, Temash and Kaveri.
In this generation, the lowest TDP chips are 17-Watt Ivy Bridges and 10-Watt Atoms / AMD Brazos. For whatever reason, AMD Brazos aren't getting used in tablets (they seem like a nice chip though...), the Clovertrail Atoms are too weak (lacks out-of-order execution, performs mediocre even compared to ARM chips, extremely poor GPU performance), and the 17-Watt Ivy Bridges use too much power compared to the 5-Watt Arm chips competition.
( Part of the reason is that Intel Atom chips are second-class citizens of Intel. They don't get access to the latest process, and were instead seen as a cheap cost-effective chip for netbooks, as opposed to a premium power-efficient but powerful chip that you'd expect in a tablet. )
With a power disadvantage, this generation of convertibles just ain't gonna cut it in terms of battery life / portability / weight. The CPUs require too much power, which require bigger batteries.
Next generation stuff from AMD / Intel however is looking quite exciting, and the potential standardization of widespread adoption of DDR4 for more power savings helps this form factor greatly.
In fact, Intel has introduced new instructions to Clovertrail Atoms and Haswell specific to new low-power states of the CPU. Intel's willing to change their instruction set to get better power savings.
True, Tablets will benefit from DDR4 (and already have the power-saving instructions Intel is only now implementing). But the real benefit of battery life stops being useful past around 8-hours or so. (With exception of e-readers and their 30+ days of battery life. Those things are amazing on vacation).
Furthermore, ARM chips are growing in a different direction in the near future... they're aiming at the server marketplace. The innovations of ARM 64-bit don't really translate to smartphones / tablets. Exciting for the server market, but I don't really see major innovations occurring for Tablet-space ARMs in the near future.
If the convertible concept survives till next year, then we'll see some real competition. In 2014, ARM chips will grow into the server market and attack Xeons / Opterons. And simultaneously, Intel / AMD solutions will have their first real salvo into the ultra-portable market.
---------
That said, the Surface needs to be made. Microsoft needs to innovate from the software side and figure out how to update Windows 8 to be more palatable for everyone. So even if Intel / AMD chips aren't ready to compete, Microsoft gotta move forward here. They're clearly falling behind in this race and need to put something out there.