The mobile computing roadmapThis week: What Intel is doing to keep the momentum in computing moving from the desktop to mobile platform. A look at Atom, WiMax, netbooks, and what to expect from the Intel Developers' Forum next week in San Francisco. Guests: senior editor Dan Ackerman,...
[Music] ^M00:00:09 >> Mr. Needleman: Hi and welcome to episode number two of the Reporter's Roundtable CNET's weekly deep dive into one major topic a week. I'm Rafe Needleman from CNET and today we're going to be talking about the mobile computing architecture and how it's evolving and changing. The big focus on what Intel is doing. We're heading into the holiday buying season and most of the laptop vendors have new lines using mostly Intel CPUs. We'll talk about these and also the Intel Developer Forum is coming to San Francisco next week. It's the biggest Intel chip conference -- one of the biggest Intel chip conferences of the year and we'll cover what might be coming up there. We have two experts this week to guide us through this. From San Diego Brooke Crothers who is the author of CNET's Circuit Blog Nanotech and also a columnist for the New York Times Bits Section. And from CNET's New York Laptop Testing Lab senior editor Dan Ackerman. Dan and Brooke welcome to the show. >> Mr. Crothers: Thanks. >> Mr. Needleman: All right. So let's get started. When I originally started putting the show together I thought I would focusing on desktop components because I build a new PC about every three years and I have to immerse myself in Intel's new lingo which is always shifting. But as I talk to you guys I realize the focus really on innovation in CPUs and the whole architecture is on the mobile side. So has innovation really moved from desktop parts to mobile parts? >> Mr. Crothers: Well, I would say just walk into Best Buy and you can get a pretty good idea immediately of what's going on. There's like 50 laptops and 10 netbooks and you look over to the side and there's a few towers. >> Mr. Needleman: Dusty. >> Mr. Crothers: Maybe 7 or 8 towers. So obviously the focus is on mobile and that's where Intel's focus is on certainly. >> Mr. Needleman: So looking at the architecture there a mobile platform is extremely different from a desktop one. You've got to focus on many different things. So what are the things that Intel and other manufactures are looking at when they're kind of architecting now for mobile primarily. >> Mr. Crothers: Well, I would say integration. >> Mr. Needleman: Uh-huh. >> Mr. Crothers: And that's really the theme at IDF. So you make 3 chips into 2 chips 2 chips into 1 chip and you move up the manufacturing process. Right now we're at 45 nanometer Intel is targeting 32 nanometer. And all of that gets you a lot of power efficiently. So for example if you had a discrete or a separate graphics chip and you move that on to the CPU and you move that up to a 32 nanometer process that gets you a lot of power savings and that's happening in the mobile space. And that's the focus for Intel's 32 nanometer is in the mobile space. Not so much -- of course they'll do desktop 32 nanometer but the real focus is on mobile. >> Mr. Needleman: Okay. >> Mr. Akerman: And you have to think about why they want to do that and part of that is so they can then go into what the end user end result of that is you can go into the store and if a machine uses less power they can put a small battery in it and still say all day computing. 8 hour computing. Whatever the big stickers they put on it. >> Mr. Needleman: Do consumers really care about batteries? I mean traditionally I've gone in you know when you go into Best Buy or Fry's or whatever you see these consumer laptops and they're gigantic. >> Mr. Akerman: Well, here's the problem. All of these laptops are essentially commodity products and for the most part they have the same exact components inside. So the only way you're able the differentiate from the competition is through it's kind of design I guess also price and then you know any kind of special features whether they're real or imagined that you can put on a big sticker on the box or put in you know the Best Buy circular. >> Mr. Needleman: So -- sorry, Brooke. >> Mr. Crothers: No go ahead. >> Mr. Needleman: So how are Intel and some of the other players that put the parts in these or make the parts for these computers how are they kind of pushing things along? I mean we've talked about the die size or process sizes are getting smaller. They're going to 32 nanometer. What else are they pushing along into these products to make them you know more mobile as opposed to just you know portable desktops but actual mobile devices? What are they pushing on the manufactures and then on to us? >> Mr. Crothers: I would say look at netbooks and look at this knew category ultra-thins. So netbooks is all about power savings, right? And it's a very small item it's a very you know -- to put it in simplistic terms a small chip and it uses very little power. So my take is that battery life is important now. It wasn't that important before but now it's really important. So you look at all the now ultra-thins that came out and I think Dan can probably validate this it's all about you know they're saying 8 hour 10 hour battery life and that's really different then before where it was just like you know giving us maybe 3 or 4 maybe 5 hours max. So you know all the technologies we just addressed this a minute ago really that enable that you know longer battery life. Because think about it. When you take a netbook or a laptop with you now you know it's going to be really crucial that you get -- if you go to an all-day conference I mean it's going to be really important that you get all-day battery life. And I don't think people expected that before but I think that expectation is there now. >> Mr. Akerman: Just very quickly on battery life. What they used to do is say, Oh, yeah, 8 hour battery all-day battery with a big asterisk next to it. And then if you read the fine print it said you know if you turn the monitor down to like 20 percent brightness and you turn off the wifi and you don't so this and you as long as you don't open any programs. Now at least they can make the same number claims but they're actually inching back towards reality. There are a lot of very fine for instance fairly inexpensive netbooks that have 6-cell batteries that don't stick out too badly from the back and they actually will run for 5 1/2, 6 hours on some strenuous stuff like video looping. >> Mr. Needleman: So let's talk about I want to learn about Atom because of course that is -- traditionally a netbook is a small computer powered by an Atom chip. What is the Atom chip? How is a different from your standard Core Duo or Core 2Duo or whatever? What makes it special? >> Mr. Crothers: Well, it's not very sophisticated. Let's put it that way. It's a pretty simple chip and if you look at the Core 2Duo and the upcoming Nehalem it's a very super scaler and all that stuff. Atom is not -- though I guess some chip people may take issue with this -- but it's not really a super scaler chip. And super scaler has been the design MO for the last ten years. So it's a really simple chip and thereby gets you the power savings. And that's the simplest way to put it. >> Mr. Akerman: So it's kind of a dumbed down chip on a modern process? >> Mr. Crothers: Yes. It's a very modern process. It's very cheap to make. And so in other words you can get a lot of Atoms on a wafer. So what Intel's thinking is we want -- here's this new business model where we have to make you know we have to make chips for cheap computers so we want to have a profit margin on this chip. So put a lot of chips on the wafer and you know therefore you get better yields. Theoretically a decent profit margin. However, I would take issue with that. And Intel is sort of -- I wrote a blog about this and actually written about this a number of times that it created a Frankenstein monster of sorts in that it's becoming too successful because Atom was originally designed for the mid-market for the mobile internet device. Think of it as a high-end smart phone and all of a sudden you know [in audible] and Acers chimed in and everyone started making netbooks all of a sudden. And Intel did not expect this. And they will tell you that they did but they didn't. >> Mr. Akerman: They became a victim of their own success in a lot of ways. >> Mr. Crothers: Yeah. Yeah. Exactly. >> Mr. Needleman: So now people in the chat room are asking did I just say that all netbooks are Atoms. Let me be clear about that. My understand was that initially people were trying to say that a netbook is an Atom powered machine although I think that has changed now and netbooks are just whatever people want to call them. And we've got netbooks and we've got netops which are basically netbooks in desktop form factors and I mean is this Atom chip line going to continue as the standard you know mobile chip, the traditional mobile chip becomes more efficient are we going to see the death of the Atom? >> Mr. Akerman: No. I don't know we will. But remember the very first netbooks actually had Celeron chips in them the Atoms didn't come until some time later. And that's still I would 90 plus percent of all netbooks. A couple of people have tried to put other chips in netbook Avia had their nano-processor I think I saw like one Samsung system ever and AMD had their Athlon Neo which they kind of pitched as a step up from the Atom. Something that was a little bit more powerful they could charge a little more for which is kind of the holy grail of everyone who is kind of stuck in this netbook moras where they can't sell a machine for more than 299 or 399 now. Which is what I meant when I said they were victims of their own success. But the problem with that Athlon Neo is and the reason why I've only ever seen it in one system is it just really wasn't that much better even though they tried to charge another 300 bucks for the final product. >> Mr. Needleman: There is getting a little far afield here but I'm worried about the netbook or the laptop market in general because of netbooks being so low cost. At 300 bucks or 250 or even if you go done even further. >> Mr. Akerman: And they will. >> Mr. Needleman: I don't understand how anybody is making any margin. >> Mr. Akerman: They're not. >> Mr. Needleman: The Windows license fee is 50 bucks itself. >> Mr. Akerman: And they might have to knock that back too which is something I think has been hinted at recently. I'll just say very briefly on that subject. A point I've tried to make repeatedly to people in the industry who may not want to hear this is that with the advent of netbooks and Atoms the consumer has finally woke up and realized that for years they've been buying too much computer. And now they understand that. And now they have a lot more power in the marketplace than when they just went into a store didn't know what they wanted and somebody said here you need to buy this $1000 system. >> Mr. Needleman: What other important things is Intel or the other manufactures doing to kind of balance the need for you know processing power and energy consumption there? >> Mr. Akerman: I think they're pushing these new consumer ULV chips which are a much different animal than the old ULV chips you used to find in those little like 12 inch we used to call them ultra-portable systems that were like you know $1,500, $2000 that now they call this knew middle ground CULV. The first examples of it however -- and I'm not sure how Brooke feels about this -- have not been super impressive to me. >> Mr. Needleman: Are these CPUs or support chips you're talking about? >> Mr. Akerman: There's are CPUs we're talking about. >> Mr. Needleman: Okay. >> Mr. Crothers: Right. So what Intel is trying to do is -- so Sean Maloney who is sort of considered to be the next -- may be the next CEO of Intel. And I talked to him about this a few months ago and he was saying you know netbooks are great and we're glad the netbook category is successful but we have this knew chip called they call it the CULV now it's called the ultra-thin category. And that's really their strategy to get back the profit margins. So I agree. It's a real challenge for Intel. The netbook has been too successful and I really don't think they can stop it. And bear in mind that at IDF Intel will be announcing a new Atom platform called Pinetrail which is the biggest probably leap for Atom in a long time because Atom you know -- did you notice this last time they upgraded Atom they went from what 1.6 gigahertz to 1.66 gigahertz. It wasn't a huge leap. So obviously they're trying to control that category so it doesn't get too high performance. But Pinetrail is going to actually put the graphics right on the CPU die so that will be interesting to see if that gets you more power efficiency and better preference. >> Mr. Akerman: But real goal there is to lock out the nVidia ion which is nVidia's graphics chip for low powered system and they're really not interested in helping you out by putting the graphics on the chip they just want to lock out the competition and make you buy their full package rather than just the CPU and then use your own you know chip set with it. >> Mr. Needleman: Let's talk about graphics for a second and extend on that. One of the things we're seeing now on mid and high-end notebooks is dual graphic systems on the same product. Lenovo and Apple both have machines and I think Sony does as well and probably some others with switchable graphic subsystems. You can go integrated which is the one that's built into the mother board or you can go the discrete graphic which is I don't know how they solder them in or socket them in or whatever. Is that the wave of the future or are consumers going to need less and less of the discrete graphics? >> Mr. Akerman: They definitely already need less of the discrete graphics. I've talked to a bunch of manufactures about this very topic this week because everyone is come around showing their holiday lines. And I think a big problem is most people what who have a machine with switchable graphics don't even realize they have it. It's almost like this instant on prelaunch operating systems. They don't even know they're there. If you look at the Apple Macbook pro there's a setting somewhere in the menu that says more battery or more performance. It doesn't actually tell you what you're doing. >> Mr. Crothers: Well, yes. That's a good question. I think and you know nVidia is not going to go away let's put it that way. nVidia is a big company and they deliver really good graphics and they have a pretty good marketing machine. It's not up to Intel's level but you know I think they make a point that for example as Dan was just talking about in the netbook space their point is if you want you know let's say laptop level performance in a netbook you should go with our ion platform. And that is of course it's on a discrete -- it's not on a dicrete chip set, excuse me, a discrete graphics chip it's actually integrated into the nVidia chip set. Just for argument sake let's say it is a discrete chip because it is a fairly high-preference chip so their argument is you know you do need a discreet chip to get real mainstream PC level performance in a netbook. So you can pay you know let's say $400 or $350 for a netbook and get mainstream PC level performance. So the point being that yeah it is necessary. I'm you do need discrete graphics level performance because Intel hasn't really delivered that yet. I mean Intel integrated graphics is let's face it. It's okay but it's not great. >> Mr. Akerman: You know it's okay if you need the computer to turn on and that's about it. >> Mr. Needleman: Oh I don't know. I use integrated graphics. I don't do -- aside from Google Earth I don't do anything that's graphics intensive. >> Mr. Akerman: But here's the test of the netbook and the reason why people are finally realizing they need something other than just the default Atom with the integrated graphics is try to go to HULU and try to watch what they call HD which is their 480p stream and 9 times out of 10 your netbook is not going to -- it's just going to stutter. >> Mr. Needleman: Okay. Point taken. And it surprises me actually that Intel is in that position. I mean AMD bought ATI shouldn't they be ahead of the game on this? You know why are they not ahead of the game here? >> Mr. Crothers: I can address that. They do have better graphics than Intel. But they really have problems with their CPU. Their mobile CPU. I'm not talking about their desktop I'm not talking about their servers. They really have a problem with it's called Turion and they have a thing called Neo. Neo is okay. It's being marketed more or less as a netbook platform but it's still pretty much Atom level preference. And Turion has never really caught fire. And Intel has just completely dominated the mobile space. And I bring this up with AMD often. Typically when you get PR from AMD it's all about desktops, yawn, and servers which are of course very important that's a very high profit margin segment for them. But you hear very little about the mobile space. These days you're hearing a little bit about Neo but you know for AMD it's a real challenge. Here's this huge huge market that's just growing leaps and bounds everyone's buying laptops and they're not a big player and they do have ATI so if you go to Best Buy you can get an AMD laptop with an ATI and it's better graphics than Intel graphics but you know the average consumer they go to the Intel solution. So that's a problem. >> Mr. Needleman: That's brand success. >> Mr. Akerman: I mean we went to some retail stores for a big back to school round up for laptops. They get all the models off the retail store shelfs and test them. And you're entirely right. You can get the AMD version a little bit cheaper than the Intel version in very similar systems but in our actual benchmark testing -- and we're talking the 499, 550, 599 systems -- there's really a big preference difference and that's enough for you to say I'm going to spend 50 bucks more and get the Intel version instead. And the problem with that Neo that Brooke mentioned and again the reason why it's really only ever been in one laptop I think a second one from HP is finally going to come out but I think it going to be a dual core Neo is instead of AMD saying hey the Atom is successful Let's make our own clone of that and try to capitalize on that they said, Let's make our own more expense clone of that and try to up-sell it when it really wasn't that much better or enough to make you pay more for it. >> Mr. Needleman: Let's move on to the last tech -- whoa I need a new chip -- the last tech topic before we get on to the IDF here and talk about mobile -- and talk about communications rather. Is this going to be the season of WiMAX? And if not when are we going to see that? >> Mr. Crothers: Well, that's a good question. I think WiMAX is a tough call because it's not as unsuccessful as people think but it's certainly not that successful yet. I think it takes time. I was reading an article on a Japanese web site I think it was yesterday. And it was about a Thinkpad that had a WiMAX you know one of the few laptops that has WiMAX built in and the writer was complaining and again this is in Japan which supposedly is you know fairly far you know relatively advanced in WiMAX infrastructure. And the whole article was about him complaining he couldn't get a good connection in downtown Tokyo. So that's a problem for WiMAX. But you know it takes time for these technologies. You have to build up the infrastructure. And I'm not going to say that you know WiMAX is toast yet. >> Mr. Needleman: Too early to bury it, huh? But we're not going to see it this year? >> Mr. Akerman: Oh, no. Definitely not. I mean what's the one test market? Is it Denver? Is that what it was? >> Mr. Needleman: I think so. >> Mr. Crothers: [inaudible]. >> Mr. Needleman: Not here of course. >> Mr. Akerman: You have the infrastructure problem. You got LTE at the same time and you have again who is the market for this? People are finally everyone finally has wifi on their laptops now. And people finally have routers in their houses and they finally have wifi in all the coffee shops and municipal wifi in the parks. Are we just going to tear that down and start again? Who's going to pay for all of this? >> Mr. Needleman: You are. That's the answer. All of us aren't we? >> Mr. Akerman: And we're also try to up-sell people to getting their own personal 3D subscription data connections at the same time. We spent years building that market up and that's finally becoming almost an acceptable thing to get when you buy a laptop and want the antenna built in. >> Mr. Needleman: We're final seeing these subsidized netbooks. You know, so -- >> Mr. Akerman: Yes. We're starting to. A handful of cases although the subsidies aren't nearly enough. You know they will sell it to you for 199 instead of 299 with a two year contract. It really should be 99 bucks, 49 bucks, free even. You get a 299 Aces 2d50 that's you know or a Dell mini 10v those are 299 netbooks. If you're going to get 60 bucks a month from me for two years just give it to me for free. >> Mr. Needleman: Yeah. No kidding. Hey, let's close with looking at the Intel Developer Forum that's coming up in San Francisco next week. I'll be down at the demo start-up conference. Brooke's coming up here to go to IDF. What do we expect to see here? How important a show is this? >> Mr. Crothers: Very important. And I'll throw the chip heads out there I'll give you some code names and you know code names are confusing but they are part and parcel of IDF. It's a big part of IDF. So what we have is we have first we have this thing called Clarksfield which is a core which Intel Nehalem type architecture. They're calling it commercially the core I3, I4, I5, I7. And so the chips that I'm going to list are mostly core series chips. The Clarksfield is a 45 nanometer quad core Nehalem mobile processor. It's the first Nehalem mobile processer and that will be a big thing at IDF. Then there's this thing Clarkdale. It gets confusing really fast here. Clarksfield. Clarkdale is a 32 nanometer Nehalem desktop chip integrating graphics on the CPU for the first time. And there's this thing call Arrandale which is a next generation sort of after Clarksfield which is a 32 nanometer Nahalem mobile chip also integrating graphics on the CPU. So as you can see integrating graphics on the CPU is a big thing for Intel. There's this thing called moristown [assumed spelling] which everyone I think it's a pretty well-known code name. It's a 32 nanometer system on a chip. And so a system on a chip is obviously just what it means everything is on one chip basically. And that's going to be Intel's -- that's 32 nanometers as I said. And that's going to Intel's vehicle to get into the smart phone states. And that's where the arm versus Intel arm being the Qualcomms, the TIs, the free scales of the world. That's Intel and I think this is going to be one of the big topics at IDF. What Intel is trying to do is they're trying to downsize into the smart phone. And that's not easy to get x86 which you know is in servers and high-end gating machines. And it's a power you know inherently a power hungry architecture. You get that in a smart phone that is no mean task. And then you have arm the arm camp trying to up-size into things called smart books. So that's an interesting thing to watch. And just a couple more things the Pinetrail thing which is the Atom chip that I mentioned. Then Laraby [assumed spelling] is Intel's first discrete graphics chip in about ten years. And that's going to be very interesting because that's going to compete with nVidia and ATI. >> Mr. Akerman: Ten years? >> Mr. Crothers: Yeah. They had a discrete graphics chip way back when that no one has ever heard of. If you know Intel you've heard of it and if you've followed the chip industry you've heard of it. But that in fat didn't pan out. So this is their first discrete being operative discrete graphics chip in about ten years I think. May be a little more than ten years. >> Mr. Needleman: Wow. All right. Brooke, thanks for dialing in. Dan, thanks as well. I tell you something this show went on a little longer than I planned but I learned a lot and I'm going to keep doing this round table just because it's incredibly educational for me and I hope everybody else is enjoying it as well. Thanks guys again for dialing in Skyping in whatever it was. Everybody thanks for listening to Reporter's Roundtable. We are life each Friday at 1:00 Pacific time at live.cnet.com. Next time we're going to have our first outside the CNET infrastructure guest, Aaron Patzer. He's the CEO of Mint or now some VP at Intuit because the company Mint is a personal finance web site was acquired by Intuit for $170 million. He'll be here in the studio to talk about what is going to happen with personal finance and Intuit and Mint. And Eleanor Mills our security reporter will also be here to help me grill Aaron. Digital City, Dan Akerman's podcast is moving from the spot on Friday to the primo spot of Mondays 3:00 Eastern time Noon Pacific. This is about technology for the urban environment. You can find Brooke Crothers on CNET on the Nanotech blog. And if you have question about this or other podcasts you can reach me at firstname.lastname@example.org. I would love to hear your feedback on the show and the show format. And that is it for Reporter's Round Table Number 2. Thanks everyone for listening. ^M00:24:19. [Music] ^M00:24:24