A year after SemiAccurate called Intel’s moronic Ultrabook spec “Shiny things for the stupid“, the world is coming around to our point of view. iSuppli has come out with the numbers to back us up too, looks like we were right all along.
Lets face it, Ultrabooks have failed because they are a fundamentally bad idea based on nothing more than Intel’s greed and fear. Intel is trying to compete against tablets and phones, and doesn’t have a clue how to do it. Ultrabooks are failing because there is no reason to buy one over a tablet or phone. Why? There are lots of reasons, but that is a separate topic. Lets just say that Intel is slavishly toeing the failed Microsoft line and not offering the consumer any real benefits. To compound the problem, they are jacking up their prices to unpalatable levels while squeezing any hope of profit from the OEMs. As we said over a year ago, there is no possibility of this debacle succeeding.
About 18 months later, iSuppli comes out with the numbers to back us up. Even with the hype, massive advertising campaigns, and second generation CPUs in place, Intel will be lucky to sell 25% of what they said. To make matters worse, the outlook is far gloomier than even iSuppli thinks, and we have good reason to believe no one other than Intel is making money. Why? The two bugaboos of the modern PC ecosystem, Intel’s greed and stupidity.
When we laughed at Intel’s theories on revolutionizing the laptop, we explained to them in great detail about why it would fail. They brushed it off and told us that fashion would conquer all. We laughed again and pointed out that it wasn’t a Macbook Air, but it did cost more, did less, and still wasn’t a Macbook. That last point is important enough to state twice. They replied that some people want fashion and not a Macbook Air. We conceded that point and the eight sales that went with it, but still questioned where the other 39,999,992 customers would come from.
Intel retorted that some people wanted or needed Windows. Parallels was brought up. They dismissed it outright calling it too complex. By now, their excuses had taken a sharp right turn from any reality SemiAccurate exists in, fighting Apple on fashion, and fighting the complexity of MacOS with Windows. To give them points, they did not try and bring up security and whatever nebulous threats McAfee made up that week.
When it came to tech, Intel pointed out that the initial Ultrabook designs were just that, a first attempt. The real versions would come out in about a year when Ivy Bridge CPUs hit the market. By then, the money they were investing in the Ultrabook Fund would drop tooling costs radically as volumes spiked, and the new “do everything” Ultrabooks would change my mind. They didn’t mention how the funds contractually exclude manufacturers from working with AMD, but that is another tangent. We laughed.
A year later, Ivy Bridge came out. The tooling ramped up, the new materials that will revolutionize everything appeared, and the second (third?) generation Ultrabooks hit the market. Initial tests showed that the battery life still was untenable, the features were still lacking, the costs were still shatteringly high, almost unchanged from a year prior, and everything that made Ultrabooks unpalatable then made Ultrabooks unpalatable now. Performance was better than before but still laughable compared to a less expensive, more functional, and slightly thicker laptop.
We brought this up to Intel, again pointing out why the category was a failure, and how it was doomed by their own greed. Intel will not address that problem because any fix would directly affect their margins, and that is one thing that will never change. Instead, they said that we should just try one, and then we would see the wonders that were Ultrabook. We asked them to supply us one. They promised to do so. And didn’t. Again and again and again.
Every time, there was a new excuse. Eventually they started insisting that we run specific OSes, run specific tests, and only do the things that they proposed or they would not supply us one. If we deviated at all, either in testing or writing, we would miss out on the wonders of Ultrabooks. We told them to get bent again. After about a dozen rounds of this, SemiAccurate gave up, but Intel is still more than happy to supply any site willing to give the category good reviews with as many as needed, prototypes, upcoming concept hardware, and anything else they need. If you hinted at an honest review however, you got lied to.
At IDF, Intel is now doing a word replace with the text on why the next generation will change everything. When Haswell CPUs come out, you will see. When the third, or is it fourth now, generation of Ultrabooks come out, they will revolutionize everything. And once again, Intel fails to address a single one of the Ultrabook’s problems, price included.
Instead, Intel is about to compound Ultrabook’s problems once again. How? Greed. The long promised $699 price point that the second third far future maybe generation Ultrabooks will possibly come close to is receding rapidly. The second generation is still no cheaper than the first, often more expensive in most cases due to feature creep. The ones that go below $1000, some even reached $800, are unusable fragile plasticky garbage that no sane person would buy. That includes the subset of customers that would not buy it over a tangibly better quality $400 notebook even though it has an Ultra- prefix.
To fix this, Intel is doing two things, and both are horrifically and comically wrong. They are vastly increasing the price of Haswell over Ivy Bridge, and are mandating feature creep. The category could not break $1000 without fatally compromising quality is now having a feature list that will bloat the BoM. Intel mandated a form factor that necessitated extraordinarily expensive materials and manufacturing techniques. The result is a product that is quantifiably less rigid than a lighter, cheaper, and less expensive laptop like the author’s Lenovo X201s.
One reason for this is Intel’s $225 or so asking price for the Ultrabook CPUs. Sure, Intel makes cheaper models, but for a category that already suffers because of poor performance and low battery life (Note: Intel bins high end CPUs on leakage as well as clock so high bins do get better battery life), that just makes those models less palatable. With Haswell, prices goes up, and the top bin gets a new tier of pricing. The top ULV Haswell has to pay for marketing of a the new CPU brand of that Intel is going to announce soon. Prices are going to end up dangerously close to $300 for Haswell. For a $699 Ultrabook, what could go wrong with that brilliant pricing plan?
In the mean time, the real problems still remain. The form factor is abjectly broken, mainly too thin for the purpose, prices are still too high, and buyers are still staying away in droves. Intel keeps claiming that this will be fixed with time, but until they find a way to update the laws of physics, don’t hold your breath. This mandatory stupidity dictates much of the expense that is not the CPU, that won’t budge either. The user gets absurd battery life, thermally limited performance that is only borderline acceptable if you run very unrealistic benchmarks no credible site will, and the necessary features that make up a laptop are precluded.
Want a realistic number of ports? Nope, VGA for a projector or monitor? Nope, too thin. Full height Ethernet? Not possible. Removable battery? Not a chance. Keyboard with actual travel? Guess why that isn’t on the cards? Any chance for expansion? Yeah right. Luckily, if you have a dongle fetish, Ultrabooks are for you, some even have a mini-VGA port that no one else does, how convenient. The entire form factor is simply dumb. None of the 20+ designs shown at IDF last month change this at all.
Once again, Intel rushes in to fix things. No, they didn’t raise the Z-height mandate. No, they didn’t lower their monopoly CPU prices. Instead, they are mandating touch screens. $75 BoM cost increases are being kind here, so a $1200 Ultrabook just became a $1350 Ultrabook. Hint: $699 is in the other direction. Luckily, Intel is also mandating NFC hardware, higher rez screens that suck more power, more expensive batteries to cope with it, gyroscopes, sensors, and lots of other, well, crap. None of this is cheap. $1350? Pipe dream.
In the end, Ultrabooks are badly broken by design and unchangeable by fiat. They are not selling, even the halved forecasts by the pessimists were 100% optimistic. The light at the end of the tunnel professed by the tame sites meekly countering iSuppli’s broadside in an attempt to salvage their advertising dollars is simply not realistic. Intel is forcing the category the wrong way in terms of price, performance, and consumer desires. Rather than fix a single one of the problems that are crippling Ultrabooks, they are trying to make 33% more on each one sold. Ultrabooks have failed, all Intel is doing now is making their death more comical. I wonder how many OEMs this will take down in the end?S|A
Latest posts by Charlie Demerjian (see all)
- SemiAccurate has Skylake-SP die shots, sizes, and more - Jul 21, 2017
- Skylake-SP has a diverged core - Jul 19, 2017
- Intel’s Purley platform architecture is a step forward - Jul 13, 2017
- Intel launches Purley aka metal Xeons - Jul 11, 2017
- AMD’s Epyc has lots of connections - Jun 26, 2017