Think PCs will drop in price? Think again, warns Intel The Golden Age of ever-decreasing PC prices is over, at least as far as Intel is concerned. Speaking to investors in London last week, the chip giant's CFO Stacy Smith boasted how the vendor had broken the pricing death spiral that has bedevilled the PC industry for most of its history. This has been achieved by a combination …
"....and able to set the prices..." The vendors only set the prices on brand new, cutting edge kit, and in the case of most people buying home desktops, laptops or home servers, they buy the latest and greatest and pay top whack out of ignorance. If you are willing to stay a step behind the bleeding edge then you can get your kit rediculously cheaply. I often advise friends buying PCs not to buy brand name desktops or laptops on current offers but to look for wholesalers shifting old stock (plenty on eBay), who will often sell you last year's latest and greatest (and often still with a manufacturer's warranty) for half the original list price. I even know some fanbois that manage to limit their spends buy buying older Apple gear. After all, apart from the really hardcore gamers, very few home users will ever use even half the power of their desktop or laptop. I even helped a friend set up a software business with half-price Xeon workstations, servers and storage, just by sifting through the wholesaler websites.
Intel may be competing effectively against Via (are they still around?) and the CPU bit of AMD, but they are being squeezed on both sides, by ARM in the low energy sector, with more and more fondleslabs replacing netbooks, and possibly ARM chips moving into the laptop and small server sector; and by Nvidia and the ex ATI bit of AMD in the high performance sector. Intel don't really have an answer to this.
That is his job. You know the one where he gets paid a load of $$$ to talk up his company.
However the persistent itch he may be feeling now is this upstart called ARM. The CPU margins with ARM are much lower. What makes the money is the volume.
If a significant number of people forgo their traditional PC upgrade cycle in favour of buying a fondleslad of some sort which will more than likely be ARM powered the overall profitability for INtel will drop.
It may also drop if a significant number of businesses shift their way of 'doing stuff' into the cloud.
Then he may need to produce bigger CPU's but far fewer of them. Intel's volumes are hit again.
My beef with Intel at the moment is this seemingly continual meddling with CPU Socketry. The roadmap recently revealed seemed to show yet more variation in the CPU socket types. That will only serve to confuse buyers even more.
That seems to be their SOP at the moment. After all who can remember how and i3-xxxx compares to an i5-xxxx or an i7-xxxxx? How many proper cores does it have? is it HT Enabled? is it VT Enabled? etc etc.
"If a significant number of people forgo their traditional PC upgrade cycle in favour of buying a fondleslad of some sort which will more than likely be ARM powered the overall profitability for INtel will drop."
I can't stand Intel holding back on giving us some cheap and decent mobile processors - but people moving to ARM? How exactly are you going to do your Excel spreadsheets and Powerpoint presentations on your fondleslab? How are you going to do all your printing? How are you going to type all those lengthy emails? How are you going to install all those annoying pieces of software only available for Windows? Will people be happy to buy something that does 30% percent of what they need? 50% percent maybe? They will still have to go back to a proper pc for the rest.
When ARM devices will universally provide a proper bios/uefi so that stuff can be installed easily on them (and not need a pc for every single reconfiguration), when they will universally provide usb host so any sort of peripheral can be connected to them (not to mention that you will need drivers for that) - maybe Intel can start to worry. In the meanwhile, they are just toys which are even more disposable than regular PC's - and need regular PC's to get anything done. So Intel profits even more.
Apple is thinking of migrating Air to ARM and even Microsoft might port Windows to the ARM architecture... I think that says it all. Besides, Linux already supports ARM and with that comes a huge collection of software (since GCC supports the ARM architecture).
Of course these processors are not comparable to the capabilities of current Intel models, although ARM is not going to surrender the embedded and low-power sector easily.
Yeah, all very funny being sarcastic. Please let me know when useful ARM stuff is actually available to buy and use - and I'll happily jump on the bandwagon. Like an ARM laptop I can install Linux on without needing another machine to load the firmware, with usb host/master ports compatible with all peripherals on the market. In the meanwhile, it looks like companies using ARM processors have no interest in producing generic computing machines. They like to make toys, which can't be upgraded, so that people can throw them away and buy new ones when they get bored with them. And people like you keep on raving how the second coming is nigh.
Having one high quality processor architecture like ARM is one thing, having the rest of the hardware ecosystem around them truly open - is a completely different thing. Maybe it will happen in 10 years time - but I think the hype is a bit ahead of itself right now. There is no need for OEM's to open up the architecture and make it truly useful, while everybody keeps on buying over priced and locked down tablets. Why would they?
A CPU is just one component of a computer/device. A high quality, open CPU doesn't make an open machine.
Apparently dual core x86 processors are about to get much cheaper http://www.reghardware.com/2011/05/26/intel_atom_cedar_trail_price_fall/
What's a "PC"? While Intel may keep good margin on server/gaming/workstation rigs, AMD Brazos & Atom will account for a lot of Personal Computing devices in future.
It's el Reg vs. el Reg. Argue!!!
Perhaps this sounds acceptable to the American business mindset? To me, it just sounds rank. Sure, as a company they will want to make profit - any company would. But to be so blatant about it?
Maybe they ought to consider the problem in the '90s was the (usually superior) clone parts driving prices down. Where are AMD now?
Oh well, there's always ARM. If they should devise a solid capable 64(+) bit server class processor, it'll certainly stir things up.
Remember that shareholders are not interested in profitability. They want increased profits. They want to believe they are putting money into something with huge returns.
If Intel's profits start reducing then the share price will go down and the share holders will want to see heads on platters.
"As a company, we're putting significant features into different parts of the market..."
Hmm, I would argue that this occurred because chip speeds aren't increasing nearly as fast anymore, so Intel *had* to differentiate its chips somehow from the previous generation and then hype this difference via marketing. Whereas before the differentiation was easy: "it's twice as fast as the last lot". Hence witness how HyperThreading was introduced precisely at the time when the P4 was red-lining (at the now-ubiquitous) max of 3Ghz, then removed for the roughly-70%-quicker Core2, then re-introduced for i7 whose speed is only about 25% faster than Core2.
Intel is in a bind because the older chips are still competitive with the new ones: any dropping of the price of the old line will undercut sales of the new range. And not because "the vendors' marketing teams had "done a great great job of cleaning up our brands"." It *is* a marketing maneouver, but I suggest only to convince us that the slowing rate of chip progress is a "good thing".
In the last few years, Smith claimed, the vendors' marketing teams had "done a great great job of cleaning up our brands".
Yeah like those i3, i5 and i7 chips. You did a really good job of cleaning that up. No wait. What socket do they go into again? 1155 or 1156?
'In the last few years, Smith claimed, the vendors' marketing teams had "done a great great job of cleaning up our brands".'
Clean up? My a%$e! There was a time when all you could buy from Intel was a single generation - with several processors, ranging within few hundred megaherts between top and bottom one. Now we have on sale at the same time stuff branded as Dual Core P, Dual Core T, Dual Core E, Core 2 Duo of several different lines, stuff branded as Pentium, two different generations of i3, i5, i7 cores, and each one of these in regular, semi-mobile and fully mobile ("um") flavours - plus Atoms. That makes it roughly five different core/architecture generations on sale at the same time. Not lines, not models - entire architectures. And that is just for the laptop and deksktop markets!
It's more like, throw every single possible combination of processor and letter soup at the market - and hope they are so confused that they will buy twice and thrice over what they need and have no idea if they get any value out of it.
Where is my souped up, cheaped down Atom, with 18 hours battery life to make netbooks true mobile phone killers? What - ah, I thought not - that doesn't fit the 50%+ margins Intel, does it now.
Clean-up? Puhleeze!
I was thinking of getting a new netbook: a little more than a year since I gave my previous one to my mum after getting an iPad, I sorta want one again. Less sexy but still a proper computer and can do certain computery stuff where the iPad lets me down...
But, looking around more than a year later - probably closer to two years since I *bought* my previous one - and the same model is still available. Same speed, same memory, same storage space, Win7Starter now instead of XP, like I care (it would get nuked with ubuntu on day 1) - and the same price.
This is not the computer industry I grew up with. And a strong whiff of ripoff, and that's coming from an inveterate Apple user (at least you still get more machine per £ over time there).
So why is it with all this increased horsepower my Windows 7 PC is no quicker at day to day tasks, like email and word processing than my Windows 2000 PC?
Its a perpetuating cycle of bloat ware requires the latest chips, and now they want to hike the prices. Whoa my tablet turns on in seconds and is lightening fast at browsing and email. Hmmm where am I going to spend my ££s?
(think uncle Albert from only fools and horses) During the 90's .....
the only real innovation after Intel moved up to the pentium processors from the 486DX processors was keeping the damn thing cool while you revved up the Mhz. oh, and they dropped MMX in there somewhere.
The next bit of "innovation" was when they moved to the slot type connection for the CPU declaring that you cant crank up the Mhz up any more, not in a socket type package anyway. of which AMD came along and pissed all over their chips and produced a faster CPU in a socket.
the new CPU's were the same as the old CPU's they just improved manufacturing to increase chip yield per wafer and hard wired them to the clock speeds they say it could run at. more power, bigger heat-sinks... more noise.
..and to be blunt, things lately have not changed that much... more cores and more cache, move memory controllers onto the cpu etc...
Its just a shame that between Intel and AMD they have beaten the market to death and there is very little chance of a new player to the game to kick them into doing some real innervation in the desktop cpu market.
Lets just hope they keep out of the mobile device market so that it can evolve to maturity and deliver some good mobile devices.
"convinced through the courts to have competitors?"
Very true.
But if Intel were to get broken up, what would the split be? Other than recent acquisitions such as Wind River, McAfee, Virtutech, etc, do they actually design or make anything significant besides AMD64 clones these days? E.g. how's WiMax doing? Suggestions welcome.
Presumably IA64 didn't get mentioned at either the UK or US investors conferences?
"This has been achieved by a combination of technology and manufacturing innovation, matched by a much tighter branding and segmentation strategy, he claimed."
Er, no. It's because the silicon components at the bottom end of the market can't fall any further. Half the price of the bottom-end netbooks is the screen. The other half is a mixture of the case, disk, power supply, commodity RAM and a dirt cheap CPU. None of them cost more than a few tens of pounds.
Moreover, most of them don't contribute to Intel's bottom line at all, so I'm not even sure that Intel are *in* this segment of the market to a meaningful extent. I suppose you could call that "branding" and "segmentation strategy" if you wanted. Others might call it losing, because if anyone ever comes up with a significantly cheaper display, prices will tumble again and Intel will wake up to discover that the only part of the PC business they are still in is "big iron".
IBM excepted, "big iron" is the market segment where old businesses go to die. The seven dwarves died there, killed by the UNIX vendors. The UNIX vendors died there, killed by Intel. And so the wheel turns...
Synopsis: "We're completely untouchable, so we can hike the prices as we please"
Market response:
- Computer/Pad manufacturers move more to ARM + OSX or Linux (as they are)
- Domestic users discover (finally - it's taken enough time!) that they don't actually need MS Office as 99%of what they do is via email/web and a phone/pad/existing PC is fine for that.
- Corporate users already moving to cloud/web based delivery of systems find even less reasons to give their users anything other that Linux + browser or OSX + browser (for the executives!)
- Cloud suppliers find they can scale these apps better with Linux or their own custom OS (off-topic - sad for me, as I really like .NET and MS Design Tools - much better than open source software - sorry, it just is - C# is fab)
At this point I would assume MS will have a little word in Intel's ear about encouraging their user base to ditch Windows + Office and suggest they read up on the 'Gerald Ratner moment' - i.e. don't tell the world you're screwing them over - they may already know it, but the sure as hell don't like you to rub their nose in it!
we live in.
A world where, really, for most people a NetBook or Tablet would suit best.
.. except that NetBooks are destroyed by Windows 7 making them slow - a NetBook with XP on the other hand will run most day-to-day software (Office 2007) without issue. But Microsoft doesn't like selling XP.
As a mildly biased owner of a 7" Samsung Android tablet I'm happy to see technology moving in that direction for the average home user. It's not likely to be killed off, unlike NetBooks. And that probably isn't great news for Intel.
If MS does not get what it wants selling the current stuff, they will move on to generate something that the market will eventually buy into.
Look at the History of the OS
Windows ME sucked, so they had to come out with XP.
Tried to get people to buy Vista, but it sucked, so extended the life of XP.
Windows 7 is good enough on new hardware, but if they can't conquer the tablet, they will
rebrand XP after making some real and cosmetic changes, possibly with something designed
to break older software ala what they did with some upgrades of office documents where they
just changed the header and refused to read the older files in the hope to convince people to
uniformly "upgrade".
...to an investor meeting have effect on one and only one market - the "share" market. As such it amount to chest thumping and boasting - it will have absolutly no impact on the PC, phone or erver market - or ant other market where real comodities are traded as opposed to the imaginary assets dished out by stocks and bonds traders...
Ken, fine analysis, but if I remember rightly the 2nd most expensive piece of a cheap laptop/netbook is the Windows licence. So there's even less wiggle room for Intel than you write about. x86 is only relevant where Window box capability is required (or maybe helpful); where that is not required, x86 compatibility isn't a big deal.
That the ever increasing use of virtualisation isn't going to impact sales? that it's not going to lower processor prices? i find that VERY hard to believe. CPU power has gone through the roof and dedicated machines hardly make use of their CPU, systems are memory hungry if anything else.
with virtualisation spreading ever further, demand for CPU's in general is going to be in decline. there's also the risk of end users becoming educated and seeing they don't need a quad core 2.7GHz processor to do their work, an older processor will do just fine as long as the rest of your system is up to spec.
i personally forsee a steady drop in prices for a good while to come.
Once CPUs hit a certain performance level, it becomes viable to run *everything* interpreted, with no native code whatsoever. This means, the CPU underneath the bonnet is irrelevant: compatibility is ensured at the Source Code level instead.
And on this side of the divide, we don't have to worry about keeping the Source Code out of users' hands. Some people will still pay for apps they could have got for nothing, if the price is less than the bother of getting it for nothing. If you force users to rely on the Internet for persistent storage (sure you can set up your own mini-cloud with DNS hacks, but again the bother involved is too much for most people; though SMEs might well be interested in buying a ready-made solution) then you can monetise that.
Well, if you can remember a bit further back than the 80486, the BBC model B had a version of BASIC that ran faster than machine code on some of its contemporaries.
But there were still plenty of machine code games for the Beeb. (And with 32K of RAM -- up to 20K of which was taken up by the bitmapped display alone -- and a processor with only about 56 instructions, there was hardly such a thing as closed source in those days.)
So, I guess the moral of the story seems to be: If you bring out a faster processor and build an interpreter on it, then somebody, somewhere will want to run native code on it.
Intel can try the strategy of raising prices to increase profits, but that won't work. People who want to pay more money for a computer don't buy PC's. They buy Macs and Macbooks (PRO). As evidenced by the Netbook phenomenon as well as the "fondle slab" phenom iPad/iPad 2, consumers aren't impressed with high tech specs enough to want to spend more money on PC's. Many aren't going to replace their PC with a new and more expensive model, in the current worldwide economy. Last year, I built myself a new Tower and it wasn't the latest, greatest intel Kit. It is a Core 2 Duo 2.3GHZ running two 9800GTs in SLI. It gets the job done for gaming, work, and content consumption. What did this beast cost me to build? $400. LESS THAN THE PRICE OF AN IPAD2. I refuse to buy overpriced PC kit just to make Intel rich. I suspect many feel the same way. If I want to spend more money on computers, I'll buy a Mac.
There must be a bottom somewhere since at some point you're just buying materials. More importantly I read this as PCs, as we currently know them, are slowly being dropped from the list of consumer items and will become business computers and even then will only exist in roles that can't be filled by the coming wave of mob-lap-tab + large screen desk dock unified device.
It is one of the things that Oracle doesn't get with respect to the Android suit and java, and that is there isn't going to be a difference in platforms so having java flavors of ME, SE & EE will be a distinction without a difference. I suppose, however, that they have to milk it now before that happens.
I always wait until the price drop before I buy anything and if Itel thinks I will shell out almost a thousand US dollars for a thin netbook they are living in dreamland I found out about this on Attack of the Show this Tuesday so Itel better keep it prices reasonable if they what me as a customer anyway if my old stuff still works why buy something new.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
