Hacker News new | ask | show | jobs
by jfpoole 5317 days ago
I'm excited about ARM processors for low-power applications but I can't see ARM replacing x86 processors (even in the consumer space) until ARM performance is comparable to x86 performance. Right now the slowest MacBook Air is 6x faster than the iPad 2; until the difference is 2x or less I just can't see companies or consumers switching to ARM.
2 comments

I think it's more a matter of the bulk of consumer "micro architecture" usage shifting from computers to "devices", a la tablets, mobile phones, cars, appliances, etc. Right now, we frame our picture of consumer micro architecture as devices that look or act like a computer, but that vision is changing rapidly. Cumulatively, these devices far, far outnumber computers.
6x the performance . . . for 8.5x the TDP. It's pretty easy to see why ARM designs are already preferable for almost anything that has a battery, and laptops are outselling desktops already. Sure, there will always be applications where single-thread performance will matter more than total performance or performance per watt for many cores - believe me, we learned that lesson at SiCortex - but those applications are not enough to sustain Intel as we know it. There's a reason they're developing MIC; without it they'd be squeezed between ARM clients and servers with 20+ ARM/MIPS/POWER/SPARC cores per chip (not even counting GPU/FPGA server plays). They need their own many-core product to compete.
I will take 6x the performance for 8.5x the TDP. My 3 year old MacBook Pro has plenty of battery life for my use case. You have to look at absolute performance numbers for the application.
The only thing I use that causes the fans on my Core 2 Duo MacBook Air to kick in is Flash, so I would definitely consider a slower CPU.

Also, note that the savings/benefits can be in ways other than more battery life e.g. price, fanless & sealed cpu, etc.

How are they mutually exclusive products?
Does it still have plenty of battery life when you're doing something besides commenting on Hacker News? Perhaps more importantly, how fast do you need a single instruction stream to be? I'm writing this on my own three-year-old MacBook Pro, but I know that its battery life is mostly contingent on the CPU remaining idle 99% of the time. For any multi-tasking workload, or even for single-task workloads if the software is written correctly, a more efficient CPU architecture with many smaller cores would offer both greater responsiveness and longer battery life even during my most computationally intensive moments. Maybe you're one of the 1% who really need higher single-thread performance, and not just because of crappy multi-core-naive software, but the market is driven by the other 99%.