Since the mainstream desktop processors of 2011 are already in development, it’s easy to make predictions about them. More cores, better voltage and frequency scaling and clock gating, improved inter-core resource sharing, and tighter CPU/GPU integration in some AMD chips. Let’s say 95% for each of those.
On mobile devices—smartphones, tablets, and so on—we’re going to see some movement toward dual-core processors, mostly ARM Cortex-A8 and (on higher-end devices) Cortex-A9. The big thing in that space for 2011, I predict, will be cheaper all-in-one chips for making low-end smartphones. Take the new BCM2157 chip, for example: it has essentially everything you need for a low-end smartphone, on a single cheap-to-produce chip. Their goal is for you to be able to buy an Android phone for less than $100, and they’ll be mass-producing and shipping phones based on that chip by mid-2011. I predict that this will only continue, and dumb phones will be gradually superseded as smart phones become cheaper and more ubiquitous.
For predictions further out, I’m betting that in five years, some of the big things will be optical interconnects, networks-on-chip, mainstream GPGPU, 3D chips, and solid state drives. SSDs are already a safe bet; the rest is more speculative.
I’d put these on PB (and give optical interconnects a low prediction since it feels like I’ve been reading about them forever), if you could operationalize them/make objective & judgeable.
Since the mainstream desktop processors of 2011 are already in development, it’s easy to make predictions about them. More cores, better voltage and frequency scaling and clock gating, improved inter-core resource sharing, and tighter CPU/GPU integration in some AMD chips. Let’s say 95% for each of those.
On mobile devices—smartphones, tablets, and so on—we’re going to see some movement toward dual-core processors, mostly ARM Cortex-A8 and (on higher-end devices) Cortex-A9. The big thing in that space for 2011, I predict, will be cheaper all-in-one chips for making low-end smartphones. Take the new BCM2157 chip, for example: it has essentially everything you need for a low-end smartphone, on a single cheap-to-produce chip. Their goal is for you to be able to buy an Android phone for less than $100, and they’ll be mass-producing and shipping phones based on that chip by mid-2011. I predict that this will only continue, and dumb phones will be gradually superseded as smart phones become cheaper and more ubiquitous.
For predictions further out, I’m betting that in five years, some of the big things will be optical interconnects, networks-on-chip, mainstream GPGPU, 3D chips, and solid state drives. SSDs are already a safe bet; the rest is more speculative.
I’d put these on PB (and give optical interconnects a low prediction since it feels like I’ve been reading about them forever), if you could operationalize them/make objective & judgeable.