[–] shawnh -1 point

My opinion is to keep the small chips for the laptops and make seperate larger forms for desktops with larger cores as in make the chips twice as large and then the cores can be speed up to like 4ghz per core or more since most programs use only one core this would significantly speed up computers. If your programs usually use one core each and you have 16 cores the other 15 are almost just wasted space. Heck expand the caching also to huge amounts so the chips themselves can store the last few programs totally and free up using ram so they are more instant. Heck increase the gpu speeds by increasing the caches for the video parts of the chips. Small is great for a laptop or if you're trying to save power but if it's just a home computer and speed is the thing you want then a few bucks on the electrical bill is like who cares my computer has instant response times and never ever lags no matter what I toss at it. I also am sick of AMD and Intel coming out with a new chip and a 5% increase in speed if that. 10% for the bottom of the line or nothing should be the motto and quit this tiny incremental crap and waiting for and extra 100mhz of speed, that is a teaser. They wonder why so many people wait ten years between new PC's, well it takes them 10 years to get fast enough to make it worth the expence, skipping the little updating and going for the prize is how you get people buying more often. I watch my core speeds on Conky and one core is 100% and the others are all like 5%. Now if they can design the chips to split the load at input and use all the cores then the extra cores are worth it but they avoid that and leave the task for the OS, a cpu should be able to do that now on all of them and it seems they probably can do that but are saving it till they can't expand in other areas so they have something later for Moores Law. I say blow Moores Law out of the water and screw the yearly stuff and make it a 50x multiple every 10 years and quit the short term wasted time and money and quit sprinting and do a marathon design/dev program.