[cromwellian]

I highly doubt that assertion, it's a handwavy extrapolation. The current GPUs have significantly higher TDP, 3 years ago, state of the art GPUs had 3 Billion transistors and power consumption over 150W.

The Playstation 3 is based on 2004 era (cut down G70 )chip designs, so you're talking mobile devices next year catching up with a chip that's almost ten years ago, based on a 90nm processor node. Today's PS4 will be based on a 28nm node, there's no way in 3 years you're going to shrink the PS4 GPU and 8GB of GDDR5 to fit into an iPhone like device and not heat up like a welding torch and drain the battery like a blackhole.

It's a common, but incorrect assertion that mobile devices are going to replace game consoles. It might replace them for casual gamers, but not hard-core gamers. I don't know anyone who wants to play CoD or Half Life 2 on a mobile device. They want to play them on a high end PC, console, or SteamBox.

[lelandbatey]

Every single time I see someone say "Things in the future aren't going to see huge performance increases like that." I smile, because the beauty of the future is that we don't know. Maybe, maybe not. Maybe in ten years we'll all have crazy brain implants made possible by some new awesome technology, and we'll be watching pinar movies as these devices render them in real time.

Probably not likely, but I'm not going to make any absolute claims about anything.

[RogerL]

To be fair, the person made no assertions, they said "I highly doubt", and couched other limits in terms of thermodynamics. I'm pretty comfortable saying that thermodynamics is not going to change in a year; at least as comfortable as saying that the sun will rise in the East next year.

More importantly to my way of thinking, the parent to both your posts is talking about company strategy. You have to base future strategy on reasonable projections. Some projections are more likely than other, and you have an n-dimensional space of Gaussian and non-Gaussian probability distributions to optimize. Fortunately, most of the likely scenarios lie near a low dimensional manifold, and we can thus say things like a cell phone won't perform like today's Kepler without running at 200C (because thermodynamics). On a website we can smile knowingly and speculate about possible futures, but company directors need to be far more prosaic and pragmatic.

[johnchristopher]

That "we can't know anything about the future" attitude is way more absolute than what the parent is saying and it doesn't foster any interesting conversations about what could happen, why, how, when and how that could relates to our present and past.