News Intel's growth halted by PC sales slump

'although still bringing in $7.4 million in revenue for the company, by far its biggest earner'

I think we may be missing a few zeroes on that one. ;-)

 

Well it makes sense, they're their own worst enemy when it comes to average systems made for general business use. If you release a decent product that basically outstrips the needs of many businesses thanks to the software simply not needing more oomph, sales are going to slow.

I'm not surprised that so many companies want in on things like the IoT and wearable computing. They're areas that are fairly far off being truly practical, which means there's loads of room for growth and innovation. As such, who wouldn't want a slice of what could potentially be a fairly stable release platform? If they can release products that are of some use to a consumer, people will continue to buy them. There will no doubt be similar slumps with tablets and phones down the line as eventually the devices catch up with what people use them for. Makes sense to keep pursuing smaller and more complicated things.

 

Don't let the economists and politicians hear you saying that jrs77.
The 20th century has been predicated on constant growth, convincing people they have to have the newest must have item.

 

Intel could make a CPU that is twice as fast and twice as power efficient as SB and aside from some laptop/tablet users, the average person still wouldn't buy it. Everyday software doesn't need anything faster than what's available. Obviously there are workstation tasks where more speed is always welcome, but most applications need nothing more than an i3. That's why I always build AMD systems for people, and these people never know the difference. I'd rather support a company that actually needs the money, as long as who I'm building the PC for isn't inconvenienced.

Though intel has FINALLY been doing a good job at lowering power consumption, continuing to lower that should be their ultimate goal. The need for higher performance has kind of plateaued, but the need for more efficiency is welcome everywhere you go. It's much easier to keep slapping in more cores and increase frequency than it is to get better performance-per-watt.

 

Intels figures are still something amd would bite your hand off for. Pc hardware sold directly to enthusiasts is a tiny % of intels business. They make there cash selling chips to Dell and Apple.

Even if they released the best cpu ever and it was 4x faster the amount of enthusiasts who buy it would still do little to affect intels bottom line.

 

Not disagreeing with you, the need has kind of plateaued and performance-per-watt is a good thing but this may not have come about from a conscious decisions Intel made but because of physics.

This article from PCPer mainly talks about the GPU process node and why we may not see a move away from 28nm until 2016, but one of the first sentences stuck out for me.

 

@Corky42
I believe intel's recent decisions are conscious choices, but there are a lot of factors contributing to that; more than what that article quote pointed out.

The main factor is competition. In the PC market, VIA is basically non-existent and AMD right now is pretty far behind, giving Intel no incentive to increase performance. In the server market, PPC, SPARC, ARM, and GPUs all have their own little niche and don't really compete with intel on the same level. The way I see it, servers these days use intel due to good performance-per-watt, virtualization, and compatibility. Intel otherwise has no major advantage in servers, but those 3 selling points are all intel really needs. Intel already knows they hold the performance crown in desktops and they clearly have no intention on improving their GPUs much. Intel can't really improve performance too much or else they put AMD out of business, which will put intel into a monopoly. I'm guessing this is one of the reasons intel has invested in so many things they have no reason to invest in, such as SSDs or microcontrollers - it's a way to hold them over in case their CPU division gets crippled by the legal system.

Anyway, since intel has no reason to invest money in designing architectural improvements, they instead spend money on new ways to produce their chips, such as FinFET. If AMD were in direct competition with Intel, I highly doubt FinFET would have been a priority, because it's yields were not great enough. But FinFET gives the impression intel is actively doing something big to help keep people interested, without gaining so much interest (or performance) that they start becoming a monopoly. The lower overclockability of FinFET was also probably intentional, because CPUs that overclock too well stay in the users' hands longer, therefore decreasing sales.

Intel is slowly walking away from AMD but I feel it is very deliberate.

This is just my speculation on the situation. I'm not saying any of it is fact.

 

I thought FinFET was a solution (of sorts) to the problems that arise from the ever decreasing transistor sizes, really strange things start happening to electrons when things get super small, IIRC quantum tunneling start to become a problem.

It seems Intel is treating Mores law almost like a commandment, there comes a time when the physics of the real world start to get in the way, while you can go smaller the trade offs outweighs the benefits.

 

I blame consoles. It used to be that PC gaming progressed so fast that you needed an upgrade every 3 or 4 years if you wanted to keep up. But nowadays my i5 sandy bridge overclocked can handle anything even though it's about 5 or 6 years old, and no modern games are hindered by it.

That and the fact that everyone wants a tablet now.