My colleagues and I have been discussing this problem for several years now -- what I refer to as the "silicon by the acre" mentality that permeates our industry.
I also agree with you that I don't see this trend changing -- not just for standard parts, but also for ASSPs and ASICs.
In an ASIC engagement model, to the customer, most IP blocks are considered essentially valueless beyond the cost of the square millimeters of silicon they occupy. Yes, the customer needs that IP in his chip, but he has an expectation that you already have this IP on the shelf and you don't charge extra NRE or royalties for putting it in his ASIC -- that the development & maintenence costs of that IP have been amortized across so many customers that his share of that cost is essentially zero.
Many years ago, this kind of "free" IP was just things like standard cells, compiled memories and clock synthesizer PLLs, but today it can include anything and everything -- microprocessors, ADCs and DACs, DDR2/3 memory interfaces -- you name it.
It is no wonder that the ASIC engagement model has dwindled over the years, in part due to the substantial R&D expenditures to develop and maintain IP blocks in the latest technologies -- just to be allowed to show up for the game -- and due to the fact that those costs actually are not amortized across so many IC developments and customers that any single customer can expect his share of that cost to be negligible.
The foundry model was designed for customers like the major PC maker you mentioned. If you're only willing to pay 10% more than the raw manufacturing cost of the silicon, then guess what? You're responsible for delivering the GDSII to the fab, testing the wafers and packaged parts all by yourself, and if anything goes wrong or doesn't work, just point your finger at the mirror!
The handset OEM gets paid by the carrier, not by the end consumer. Just because you can get a "free" phone at any carrier when you sign a 2-year service contract does not mean that the BOM cost for that phone was $0.00 and that the OEM made zero profit.
Likewise, if kdboyce's estimate is correct, then LG has about $69 of cost sunk into each of these new Android handsets. Is LG wrapping a $20 bill around each handset it ships? Of course not!
Smartphones should be priced lower. My HTC incredible cost me $150 as an upgrade, but Verizon wrote it as $599 with a discount. iSupply prices the phone at about $160 in parts so there is clearly some dishonesty going on here. Most phones sell for $199 and lock you into a $40 voice plan and a $30 data plan at a minimum. This made some sense when they actually cost 600 to make, but now they don't and the customer does not see any of that benefit. If a customer does not see any benefits from a drastic decrease in costs, then there must be severe lack of competition, i.e. collusion.
One of the worst things that happened recently was the failed experiment of the Nexus One. Had this phone been offered at a reasonable price instead of at the fake price Google's partners are selling the devices, then maybe there would be more of a market for no contract services such as Metro PCS. Instead the industry can just claim that the customer does not want a contract free smartphone.
I just had the pleasure of helping two relatives who just bought the LG Ally phone from Costco (they have it prices at $0.01 with 2-year contract). I was surprise at how well it ran. The Droid operating system ran very well, even though it was a slower processor than current state-of-the art. Apps all ran a bit slower but than my HTC Incredible, but if you did not have a comparison you would have never known. The hardware is designed for a slower price, smaller screen, slower processor, lower res camera, much smaller memory storage, etc. but ran pretty well. It has the slideout QWERTY keyboard, which many older generation first-time non-tech people seem to favor, but also the touchscreen keyboard. Software was also a bit lacking in comparison, but OK for this phone. One unusual software problem was the lack of ability to change the number of rings before the caller went to voicemail. Only way to change it is to bring it to the Verizon store... Not so great....
I believe we are starting to see an attempt to lure in those non-smartphone users away from their regular phones by competitively pricing a lower-end phone that is no longer state-of-the art. Most new smartphone users will be thrilled with this phone, if for nothing else the ability to have apps through the Google Marketplace and not have to buy am iPhone.
Cell service pricing is little less than predatory. Consider Big V. A phone can be "free". Like lunch, I suppose. But what goes on? The cost of the glorified walkie talkie is amortized in the various monthly costs, and locked-in by the outrageous early termination fees. How else could Big V offer a "free" phone every two years?
OK. That said, the best strategy is to replace your handset every two years. Batteries strangely seem to last about that long, and it's not trivial to find new ones that are actually new, and not "reconditioned" (what does that mean, anyway) or just old stock.
I used to use a Palm Pilot extensively. I picked up a Palm Centro on V, and combined two functions. Wasn't as good either a dedicated phone and dedicated Palm Pilot, but I managed.
Now comes the killer. The Centro is 2 years down the road. There's no replacement, at any cost, that doesn't require either a $9.99 or $29.99 "data plan". I don't want a "data plan", but tough.
Why, oh why, can't I just pay as I go? Despite the reasonably decent connectivity that I have had with Big V, I may just dump it all and go with some cheap-o outfit.
$49 for an Android, indeed.
Aha, so LG finally thinks it needs to be in the Android game now dominated by HTC and Samsung. They are gambling by throwaway a few million $ and see how it pans out. The thing about Android phones is that they need very good response time to be appreciated and if all LG is doing is dumping android stack on cheap hardware without caring for end user experience then they are harming themselves.
That price point sounds more like a play to try to get some market share in an already very crowded market than anything else. As Frank pointed out, the price differential to the consumer over two years is almost inconsequential. However, if the cellphone manufacturer isn't sharing in that, where does that leave them?
In reply to goafrit and KB3001:
I was once told, in all seriousness, by the procurement manager of a major PC maker that they would only pay 10% over the cost to manufacture an IC. They had spent considerable effort to understand IC manufacturing cost per die....and figured they could enforce their rule just because of the market position they held.
I told them they would not have a supplier from our company, and probably not many others with that rule in place. It seems they want $$$ for their engineering and R&D but were not willing to pay for their supplier's engineering and R&D. They have since been acquired.
Unfortunately, this type of attitude by OEM's, as well as the attitude and necessity to keep IC capacity as full as possible, leads suppliers to drive IC costs down. And...major OEM's always insist on second sources for key devices, which are then played off each other to drive prices down further.
I don't really see this trend changing for commodity, or even some application specific parts, used in high volume. Part of the reason for it is that non-US IC makers who will work with smaller margins are eager to try and fill any supply gaps created by other IC makers trying to raise margins.
I have even know some off shore suppliers to deliberately buy into the market with low prices in the beginning to carve out a niche product area from a US maker, who shortly afterward, abandons it.
This is not to say you cannot raise margins and keep customers. You can..over time, but only with a continually cultivated, close, and necessary relationship with your customer base that includes your best in engineering and R&D, as well as volume, trouble free manufacturing and delivery. The real problem being solved is ultimately the customer problem, and in doing so, the supplier problem gets solved.
Those companies that do this well will survive.
My Mom the Radio Star Max MaxfieldPost a comment I've said it before and I'll say it again -- it's a funny old world when you come to think about it. Last Friday lunchtime, for example, I received an email from Tim Levell, the editor for ...
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...