Why is Intel so afraid of people asking questions about it’s new ‘Ultrabook’ spec? Intel is going very far out of its way to keep anyone even slightly critical of their new marketing co-op fund vehicles in the dark.
First a little background on Intel, Ultrabooks and SemiAccurate. When the first Ultrabook concept came out, we thought it sounded interesting until we saw the final products. They forced a form factor on users in the name of fashion that was, well, far too compromised to real world use. Why? Size.
The Ultrabook spec calls for 20mm or thinner body, and a maximum 17W CPU. Fair enough until you look at what that means. First is that if you want to have any sort of ‘style’ to it, that would be rounded or bevelled edges, and the currently ‘in’ wedge profile, your space for ports is rather limited. No, it is utterly hosed, you are lucky to get the bare necessities in.
Once you back out a bit of that height for the screen, they are thin but not zero height, you have a lot less than 20mm to shove ports in to. VGA, DVI, and even ethernet is right out, but ‘mini’ versions of some are possible. Space for USB, audio in and out, and card readers is at a premium, so you end up simply losing out on needed ports.
If an Ultrabook has three USB ports, it is nearly a headline event, two is much more common, and one is not unheard of. Most of the other scant few ports you get those ‘mini’ variants, or worse yet, proprietary cables with a dongle you have to carry for a week or two until it gets left somewhere. Even if you don’t lose them, it is mandatory bag weight that somehow is not included in the mass calculations of the machine itself.
That same lack of height also hammers battery capacity, there isn’t much volume, period, so you can’t stuff a real battery in. Since the slim profile necessitates a sealed case, a replaceable battery isn’t possible for an Ultrabook. Pity the poor exec with one of these toys on a long flight. To make capacity somewhat tolerable, OEMs have to use very expensive materials to get the meagre numbers they do.
The same problems hits the cases, to make them thin, light, AND strong enough not to self-destruct before the shrink-wrap is removed, it takes a lot of engineering work and expensive materials. Torsional rigidity is tough to engineer, and stiffness in far less than 20mm of height on a 13.3″ chassis adds cost. Lots of cost.
To get to the rather mythical $600 price point Intel still comically insists is possible, they are throwing large wads of money at OEMs. There is a fund of several hundred million dollars to grease the wheels of component manufacturers and effectively subsidize the prices that OEMs pay for parts. Not coincidentally, anyone using this fund is excluded from selling those parts to Intel’s rivals.
Even with this, the price of Ultrabooks are nowhere near the $600 promised last year, not even close. Why? The cheapest of the two iSomethingmeaningless CPUs with a U after them, the ULV parts, is $225. Stop and think about that for a moment, if the CPU itself is more than 1/3rd of the price Intel is calling for the entire machine to be, anyone think $600 is achievable? That doesn’t leave a lot for the screen, SSD, RAM, or physical construction.
The few Ultrabooks that have dipped below the $999 mark, proudly bearing the “Less than $1000!” banners on press releases, are simply garbage for quality. They eschew SSDs for magnetic HDs, slowing the already pokey machines to a crawl. The plastic on them would wistfully look up to the quality of a 1980s Detroit sedan interior, and you just know it won’t last. But it is inexpensive.
No, not $600, not even close, but a bit less than $999. That is with hundreds of Intel kickback funding rolled in. It is farcical, but Intel still insists that a $225 CPU is compatible with an ultra-slim $600 notebook. You couldn’t clue these people in with a baseball bat, they will tell you it is possible with a straight face while denying the triple digit kickbacks at the same time. Riiiiight…..
Today, Intel did something even funnier, they sent out 13″ Ivy Bridge notebooks with a 240GB SSD, a 1600 * 900 rez panel, and that aforementioned $225 CPU. Why is this funny? Because Intel told reviewers that similarly specced Ultrabooks would cost around “$1,000-1,100”. Anyone think a $334 SSD coupled with a $225 CPU is compatible with a $1000-1100 computer? We haven’t even gotten to the expensive screen yet, or the advanced chemistry lithium battery……
That isn’t the funny part though, it does get better. How? No reviewer that we can find actually questioned this little fantasy. No, big fantasy. Prices have come down a bit since the first Ultrabooks last year, but a quick glance around the web shows lesser specced models are still pushing $1500. For quick reference, $1500 is on the wrong side of $1000-1100 from $600. Intel doesn’t seem to understand that kind of math though, nor do most reviewers.
Which brings us to the opening point of the story, Intel is scared silly that people will point this kind of thing out. How scared? If you are even mildly critical of their new marketing schemes, you are cut out. No samples, no briefings, no nothing, just silence. While this behavior is positively mature compared to some of their recent PR antics, it is still stupid and childish. Then again, they threw a senior exec under a bus rather than admit an error, so why expect rationality?
When SemiAccurate first wrote articles critical of the Ultrabook spec, and pointing out how the self-professed 40% consumer marketshare and $600 price points THIS YEAR were laughable, Intel tried to convince us we were wrong. They failed, mainly because they couldn’t come close to backing up their talking points.
Once their rhetoric was fully put to rest, Intel tried the thing that they should have in the first place, promised us a chance to actually use an Ultrabook to test our theories. So they promised us one. Then another, and another. But never actually did anything about it. Other than promise one the next time something or other was going to happen. Rinse and repeat.
The closest we got to one was a specific model late in 2011, and that went poof when the last of Intel’s little hidden ‘features’ came to light. Ultrabooks meet a lot of arbitrary specs for boot, sleep, wake up, and similar metrics. Actually, they don’t really meet those specs, but if you use an Ultrabook set up with their magic drivers, and their Windows builds, and run the tests they want you to, the numbers look really good.
If you venture off the magic testing path, Ultrabooks fall flat because of the horrible technical compromises needed to fit the form. All the magic numbers that are in the sales material are unachievable in the real world, and we don’t just mean the $600 price. Or is it $1000-1100 now?
Intel made the mistake of asking what we wanted to do with the impending beast. We replied that we would do what we usually do, wipe the hard drive, install Linux, and use it for a while. See how the battery holds up in the real world, then run a few canned benchmarks against anything we felt was a relevant competitor.
You’ll never guess what happens next? No, not just that the offer was rescinded, we were flat out told that we would only get one if we used it with Windows, and tested it with Windows. Any guesses as to why? Can you imagine if Intel were to have their toys tested on a level playing field by people willing to question their extremely stretched version of the facts?
So, the last promised Ultrabook from Intel did what the previous ones did, sank beneath the waves without so much as a whisper. Not that we care about another prototype part to fill our shelves, but we did want to test their claims. Since that time, we have had more than enough hands-on time with real world Ultrabooks to know the Intel claims are completely laughable.
Because we actually question them, and won’t do exactly as we are told to do, we don’t even get briefed on upcoming CPUs. How mature. If you are wondering why Intel only gets positive reviews, now you know. They are actively shutting out press who don’t play ball, and that is sad to watch. Then again, Nvidia tried this to great effect too.S|A
Latest posts by Charlie Demerjian (see all)
- Spin Transfer Technologies talks about their ST-MRAMs - Sep 27, 2016
- ARM adds CMN-600 interconnect and DMC-620 memory controllers - Sep 27, 2016
- Globalfoundries 7nm process isn’t even close to the name - Sep 26, 2016
- ARM upgrades realtime offerings to v8-R and adds Cortex-R52 - Sep 21, 2016
- Everspin and Globalfoundries team up for embedded ST-MRAM - Sep 15, 2016