4 ms·

For those that aren't familiar, SemiAccurate is more of a tabloid rumors site than a hardcore technology site. Sort of like The Sun for silicon. They love to b

For those that aren't familiar, SemiAccurate is more of a tabloid rumors site than a hardcore technology site. Sort of like The Sun for silicon. They love to blow failures well out of proportion. BumpGate (nVidia's packaging recall problem), Fermi (nVidia's late and hot GPGPU), Tegra (nVidia's... are we seeing a pattern?) Sort of like the drunk and bitter (but frankly more knowledgeable) uncle of ExtremeTech. I'm a regular reader, and enjoy it, so this isn't an attack... I'm just recommending, along with the articles, an additional dose of a large grain of salt. Especially if it's a nVidia article.

Toshio 14y ago
SemiAccurate is very deeply embedded in the semiconductor industry and in the OEMs. Charlie Demerjian has a track record of relaying privately held opinions of executives in those industries, stuff that would be a PR disaster for those organizations should their CEOs decide not to shut up. I personally trust Charlie's analyses.
- justin66 14y ago
  Yeesh. I'm glad you mentioned his name, I wouldn't have noticed it. My sole exposure to his writing has been the occasional anti-Nvidia tirade. The stuff he wrote for the Inquirer was at a "have you ever seen a commie drink a glass of water?" level of nuttiness. It's impressive that he got a whole article out of "I asked them for a review unit lots of times and they kept saying no". I didn't see anything insightful there at all.
  - Toshio 14y ago
    One of the article's key insights was this: for any *book to be successful, battery life is the killer feature. Actually, portability is the real killer feature, but battery life is a good proxy for that. Intel chooses not to focus on that in their specs, instead mandating all sorts of bells and whistles to compensate. Not a winning strategy.
ChuckMcM 14y ago
I find it interesting that the Macbook air (Intel based) isn't considered an 'ultrabook' by them. I have felt that Intel's push for Ultrabooks was to get people making them that couldn't push them around like Apple can on price (and hence margin), except that makes them uncompetitive. I mentioned this in a talk on Saturday as well, the real revolution here is that ARM gets away with not being 30% of the COGs of the system and that changes things in the tablet/phone market. Only a matter of time before the laptop market goes that way as well.
- Legion 14y ago
  > I find it interesting that the Macbook air (Intel based) isn't considered an 'ultrabook' by them That's because they're talking about Ultrabooks, capital "U", a spec for PC laptops (and an Intel trademarked name).
  - ChuckMcM 14y ago
    I went back and re-read it, they do consider the Macbook Air to be an exemplar of what an Ultrabook should be. See this quote: "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." This is the story as I understand it: Intel wants other people to make a Macbook Air functional clone but also want their "traditional" huge chunk of the COGS. In the classic sense they have priced themselves out of the market.
mtgx 14y ago
Yes, it's a pretty biased article, but I don't mind. You know why? Because he's been pretty spot on. Anyone who's been watching Intel knows how misleading Intel can be (before anyone accuses me of being an AMD fanboi or something, I own a quad core i7 SNB laptop), and it's obvious ultrabooks will not get to the magical $700 price-point anytime soon (I'm thinking 3-4 years here). Why not? Well, because while Intel keeps promising that, they have no intention of being themselves part of that price reduction. They have no intention of reducing their prices. So they want to put the whole responsibility on manufacturers. Unfortunately for them, manufacturers have to worry about more than just CPU's to be competitive with ultrabooks. They have to worry about having a quality and high-resolution display. They have to worry about having a fast and large amount SSD, and so on. Why is it that the laptop/PC market has been considered non-innovative in the past compared to the smartphone/tablet market? Because all the focus has been on CPU's and that has helped Intel. But now people are starting to prioritize more than the CPU in their laptops. In fact the CPU may not even be their #1 or #2 priority anymore. A good display or a good SSD or high quality materials might be. This sucks for Intel, as the their processors are becoming more than "good enough". ARM devices are "good enough" for most people in terms of performance (iPad's success is proof of that), and an Intel chip that is 10x more expensive will soon become irrelevant for most people. That price gap will just not be worth it to them.
- ajross 14y ago
  Intel's CPU prices (like all semiconductor products) have been going steadily downward for... basically forever. I don't have a clue what you're talking about. I guess you're saying that Intel gets a higher margin for its parts per-die-area than fabless SoC vendors do, which is undeniably true. But generally that's considered a good thing for the company, not a bad one. Intel has competing parts in the market for almost everything they sell, and still manages to make a product that people are willing to pay significantly more money for. It won't last forever, obviously. Eventually the desktop market will dissolve and servers and mobile SoCs will be all that's left. And Intel certainly seems to be aware of this, c.f. all the marketing effort being pushed into the various low-margin Atom lines.
  - 14y ago

Toshio 14y ago

SemiAccurate is very deeply embedded in the semiconductor industry and in the OEMs. Charlie Demerjian has a track record of relaying privately held opinions of executives in those industries, stuff that would be a PR disaster for those organizations should their CEOs decide not to shut up. I personally trust Charlie's analyses.

justin66 14y ago
Yeesh. I'm glad you mentioned his name, I wouldn't have noticed it. My sole exposure to his writing has been the occasional anti-Nvidia tirade. The stuff he wrote for the Inquirer was at a "have you ever seen a commie drink a glass of water?" level of nuttiness. It's impressive that he got a whole article out of "I asked them for a review unit lots of times and they kept saying no". I didn't see anything insightful there at all.
- Toshio 14y ago
  One of the article's key insights was this: for any *book to be successful, battery life is the killer feature. Actually, portability is the real killer feature, but battery life is a good proxy for that. Intel chooses not to focus on that in their specs, instead mandating all sorts of bells and whistles to compensate. Not a winning strategy.

ChuckMcM 14y ago

I find it interesting that the Macbook air (Intel based) isn't considered an 'ultrabook' by them. I have felt that Intel's push for Ultrabooks was to get people making them that couldn't push them around like Apple can on price (and hence margin), except that makes them uncompetitive. I mentioned this in a talk on Saturday as well, the real revolution here is that ARM gets away with not being 30% of the COGs of the system and that changes things in the tablet/phone market. Only a matter of time before the laptop market goes that way as well.

Legion 14y ago
> I find it interesting that the Macbook air (Intel based) isn't considered an 'ultrabook' by them That's because they're talking about Ultrabooks, capital "U", a spec for PC laptops (and an Intel trademarked name).
- ChuckMcM 14y ago
  I went back and re-read it, they do consider the Macbook Air to be an exemplar of what an Ultrabook should be. See this quote: "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." This is the story as I understand it: Intel wants other people to make a Macbook Air functional clone but also want their "traditional" huge chunk of the COGS. In the classic sense they have priced themselves out of the market.

mtgx 14y ago

Yes, it's a pretty biased article, but I don't mind. You know why? Because he's been pretty spot on. Anyone who's been watching Intel knows how misleading Intel can be (before anyone accuses me of being an AMD fanboi or something, I own a quad core i7 SNB laptop), and it's obvious ultrabooks will not get to the magical $700 price-point anytime soon (I'm thinking 3-4 years here). Why not? Well, because while Intel keeps promising that, they have no intention of being themselves part of that price reduction. They have no intention of reducing their prices. So they want to put the whole responsibility on manufacturers. Unfortunately for them, manufacturers have to worry about more than just CPU's to be competitive with ultrabooks. They have to worry about having a quality and high-resolution display. They have to worry about having a fast and large amount SSD, and so on. Why is it that the laptop/PC market has been considered non-innovative in the past compared to the smartphone/tablet market? Because all the focus has been on CPU's and that has helped Intel. But now people are starting to prioritize more than the CPU in their laptops. In fact the CPU may not even be their #1 or #2 priority anymore. A good display or a good SSD or high quality materials might be. This sucks for Intel, as the their processors are becoming more than "good enough". ARM devices are "good enough" for most people in terms of performance (iPad's success is proof of that), and an Intel chip that is 10x more expensive will soon become irrelevant for most people. That price gap will just not be worth it to them.

ajross 14y ago
Intel's CPU prices (like all semiconductor products) have been going steadily downward for... basically forever. I don't have a clue what you're talking about. I guess you're saying that Intel gets a higher margin for its parts per-die-area than fabless SoC vendors do, which is undeniably true. But generally that's considered a good thing for the company, not a bad one. Intel has competing parts in the market for almost everything they sell, and still manages to make a product that people are willing to pay significantly more money for. It won't last forever, obviously. Eventually the desktop market will dissolve and servers and mobile SoCs will be all that's left. And Intel certainly seems to be aware of this, c.f. all the marketing effort being pushed into the various low-margin Atom lines.
- 14y ago