Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×
Intel

Intel Spills Beans On Santa Rosa Notebook Platform 96

Steve Kerrison writes "From the Intel Developer Forum in Beijing comes news of the successor to the Napa notebook platform. Santa Rosa, which will head up Intel's notebook technology line-up until 2H 2008, beefs up almost everything seen in Napa, from graphics to WiFi. 'Santa Rosa carries Robson Technology, now known as Intel Turbo Memory, the flash-based disc-caching system that speeds up loading times of frequently-used data. Santa Rosa is an obvious continuation of the Centrino series. There will also be another Santa Rosa Centrino variant — Pro — that covers the business features found on Intel's Q-series chipsets, namely vPro.' Intel's Core2 mobile processors remain a key part of the platform, as you'd expect, with 45nm 'Penryn' CPUs making their way into the Santa Rosa refresh in 2008."
This discussion has been archived. No new comments can be posted.

Intel Spills Beans On Santa Rosa Notebook Platform

Comments Filter:
  • Krikey...

    Santa Rosa?
    Robson Technology?
    Intel Turbo Memory?
    Q-series?
    vPro?
    Penryn?

    My brain can't take any more buzz.

    • by Foamy ( 29271 ) on Monday April 16, 2007 @06:49PM (#18758627)
      Apple versions

      Santa Rosa eXtreme
      Robson eXtreme Technology
      Intel eXtreme Turbo Memory
      Q-series eXtreme
      vPro eXtreme
      Penryn Xtreme Core 2
      • Re: (Score:2, Informative)

        Ignoring the fact that those are codenames, how is that at all consistent with with apple's branding conventions? They only "extreme" products are the Airport Extreme (not eXtreme) and Quartz Extreme. There is also no evidence whatsoever that Apple will be extending the use of the "Extreme" modifier to many other products. Right now, apple- branding is on the rise, and i- and Mac- branding make up most of the rest of their product line. Have you been too busy trolling to notice the names of the products you
    • I see the Turbo function makes a comeback, I had this on my P166 back in 1998, good technology never dies. I ruled back then, until Intel came out with the P3 and made the internet even faster.
      • Last I knew, the Old "turbo" function slowed the system down to the speed of a classic IBM PC, for those programs and games which couldn't handle the speed of an up-to-date processor.

        Boy, did that help for those games that timed things in instruction cycles rather than a high-res clock.
        • by DanJ_UK ( 980165 ) *
          I recall the turbo button on my goldstar 486 clocking the little beauty from 33mhz upto 66mhz
    • by Sczi ( 1030288 )
      Amen, and bravo to someone for saying it. I am sick-to-damn-death of codenames. Sick. To. Damn. Death. Stopit.

      Maybe it's my geeky math side, but version numbers have always just sort of made sense to me. Is this an increase from the Pentium 4.3.1 to the 4.3.2, or is it from 4.3.1 to 4.4.0? Even if you have no idea what the numbers mean, you can look at it, and it *means* something. If the middle number increases, you need a new motherboard or whatever.

      Trying to discuss anything today is just retarded:
      "bro,
      • by Molochi ( 555357 )
        This is assuming version numbers represent some transparent indication of progress of performance in chip design and that the manufacturer wants to convey that information. The purpose (at least towards those outside the hallowed halls of the engineers doing the work) of a code name is to intentionally promote a level of obscurity. That outsiders catalog and rank these names is contraproductive to this purpose. Manufacturers produce brand names for public consumption to indicate what they want you to believ
    • by zuhaib ( 1089275 )
      Napa... Then Santa Rosa... At this rate, they will be at Cupertino is no time! Also, i wonder how the Berkeley chip will run, or will it require a ounce of pot every 6 hours Ok i am done =D
  • by JanusFury ( 452699 ) <kevin.gadd@gmai l . c om> on Monday April 16, 2007 @06:32PM (#18758353) Homepage Journal
    This must be rough for Intel. Spilling beans on a computer is bad enough (have you ever TRIED getting beans out of electronics? It's a nightmare!), but spilling them on a development prototype? Somebody had to get fired for this debacle...
  • obsolete technology? (Score:3, Interesting)

    by getNewNickName ( 980625 ) on Monday April 16, 2007 @06:33PM (#18758381)
    Just a question, but doesn't flash-based HDs make this an obsolete technology already?
    • by lmnfrs ( 829146 )
      You didn't say what you're talking about, but I'll assume you're referring to Robson. Flash-based hard drives don't really compare because of the interface. I'm don't actually know how fast PCI-E is, but I know it's much faster than the current SATA (300Mbits/s).
      • Isn't the current SATA (2) actually 1.5Gbits/sec? Still not as fast as PCI-E, though.
        • by lmnfrs ( 829146 )

          Okay, I don't know where I pulled my number out of.. SATA2 is actually 3Gbits/sec.

          And from Ars Technica:

          "Each x1 lane of PCIe 1.1 offers a 250MB/s transfer rate, which puts a x16 link like the ones that host some GPUs at 4GB/s. The new PCIe 2.0 spec will double the per-lane speed to 500MB/s, boosting a x16 link to 8GB/s. This puts a x16 PCIe 2.0 link in the same ballpark, at least in terms of peak transfer speed, as a frontside bus based on the HyperTransport 1.0 spec (12.6GB/s aggregate bandwidth over tw

          • by cheezedawg ( 413482 ) on Monday April 16, 2007 @07:50PM (#18759655) Journal
            To clear things up- both SATA and PCIe employ 8b/10b encoding. Each byte is trasmitted as a 10 bit symbol. So 3.0Gb/sec = 300MB/sec.

            First Gen SATA = 1.5Gb/sec = 150 MB/sec
            Gen 2 SATA = 3.0Gb/sec = 300 MB/sec

            First Gen PCIe = 2.5Gb/sec bidirectional per lane, so x1 = 250 MB/sec full duplex (marketing types sometimes say this is 500MB/sec)
            Gen 2 PCIe = 5.0Gb/sec bidirectional per lane, x1 = 500 MB/sec full duplex.

            I guess the big difference here is that PCIe is full duplex, SATA is not.
            • To clear things up- both SATA and PCIe employ 8b/10b encoding. Each byte is trasmitted as a 10 bit symbol. So 3.0Gb/sec = 300MB/sec.

              First Gen PCIe = 2.5Gb/sec bidirectional per lane, so x1 = 250 MB/sec full duplex (marketing types sometimes say this is 500MB/sec) Gen 2 PCIe = 5.0Gb/sec bidirectional per lane, x1 = 500 MB/sec full duplex.

              You're sort of right, except 2.5 Gbit/s with 8b/10b encoding results in 2 Gbit/s of data ~ 238 MB/s
              So a full duplex PCI-E x1 lane is theoretically capable of about 477 MB/s
              Your SATA rates become 1.2 and 2.4 Gbit/s, 143 MB/s and 286 MB/s
              Mind you those are apples to oranges comparisons, there isn't much sense in comparing PCI-E to SATA. The narrowest PCI-E link in your system might be a whole four lanes wide anyway.
              You'd have to do a lot more research to find more realistic peak data rates, I'm

    • That would be good use for the extra pci-e x1 slots.
    • Just a question, but doesn't flash-based HDs make this an obsolete technology already?

      When solid-state hard drives catch up to magnetic platter based hard drives in total capacity, price-per-GB, and expected life expectancy, then Robson technology might be obsolete. When solid-state hard drives become available for Fujitsu notebooks, the 16GB drive will be a $700 option and the 32GB drive will be a $1200 option [arstechnica.com]. A 64GB drive has been announced by Samsung [arstechnica.com], but who knows how much that will cost?

      Last time I checked, magnetic platter based notebook hard drives have reached 250GB. Some big Del

    • I suppose it would be obsolete if you used Robson with a flash hard drive. The problem is that flash is still 10x more expensive than a hard drive for the same capacity. I can get a 200GB hard drive for under $200 now, someone pointed a link to a 32GB flash drive for $500. If you never need to use a lot of media, then maybe the flash drive will work for you, but the capacity is such a huge step back that it's probably not worth it except for a very few. Even for OS + applications on the flash, with othe
  • System Memory (Score:4, Insightful)

    by Smitty825 ( 114634 ) on Monday April 16, 2007 @06:43PM (#18758543) Homepage Journal
    One of the current limitations (IMHO) of the current Napa based systems is the fact that system memory is limited to 3GB. (Well, I guess you can install more memory, but the memory beyond 3GB isn't used) I've been following the news on the Santa Rosa systems, and I haven't seen any updates if they are going to remove this limitation, especially considering the Core2 processors are all 64-bit...
    • Re: (Score:1, Insightful)

      by lmnfrs ( 829146 )

      One of the current limitations (IMHO) of the current Napa based systems is the fact that system memory is limited to 3GB.

      I have been told that is an OS limitation, it has nothing to do with the platform (i.e. Vista 64 will recognize all 4GB).

      • by dhovis ( 303725 ) *

        No, MacBooks and MacBook Pros suffer from this chipset limitation (it is actually about 3.3GB or so), and MacOS X is 64bit capable. Mac Pros can handle up to 16GB of memory.

        I'm pretty sure Santa Rosa lifts that particular restriction, but I don't have a source for that.

      • Re:System Memory (Score:4, Informative)

        by shawnce ( 146129 ) on Monday April 16, 2007 @07:28PM (#18759273) Homepage
        No it is a limit of the chipset. The Napa chipset only support 32 bit physical addressing and a portion of that physical addressing range (starting from the largest address on down) is reserved for interfacing with the south bridge, PCIe buses, integrated graphics (if being used), etc. How much gets reserved is under software control but IIRC at least 256 MiB if not 512 MiB must be reserved. Also if certain hardware features are being used more must be reserved.

        This is all outlined in the developer docs for the Intel® 975X Express Chipset [intel.com]
    • by Idbar ( 1034346 )
      In the pictures they all seem to run Vista, so.. I guess they are not that bad, but who knows how long did they take to boot up.
  • by Anonymous Coward
    This all sounds pretty ho hum. What it looks like is pushing performance harder and harder. More performance almost always means more battery consumption. So, we'll get really powerful laptops whose batteries last at least a couple of hours. How about a radical idea; a computer that is just powerful enough to do spreadsheets and word processing with a battery that lasts long enough to fly from New York to LA.
    • by sykodoc ( 763810 )
      errr, you mean an iBook? That's already been and gone, thanks.
      • Under Linux I got my old Clamshell iBook's power usage down to 8W, and that's without CPU frequency scaling.

        For that matter, I couldn't get it to go ABOVE 15W.

        With one of those new high-capacity batteries that worked out to more than 12 hours of battery life. 8+ under normal use.

        Not too shabby if you ask me.

        Shame I broke my battery. >_<

        -:sigma.SB

    • Battery life should be better with the flash support added to the HD. I don't have the figures, but I'd guess the HD and the LCD are the main battery draws on average. I expect battery life to be the same if not a little better overall.
      • No, CPU is probably #1...at least when its running fast. It probably puts out more heat than LCD and HDD total energy combined.
    • There was once a really nice device called the HP 200LX. That had an 8MHz processor and could do most useful functions without eyecandy. Battery life was weeks with typical usage. http://en.wikipedia.org/wiki/HP_200Lx [wikipedia.org]

      There was also a cool device called the Psion 7 that could do most useful stuff and also had a good battery life.

      Sure, both those devices are clunkers by today's standards but by using modern parts they could be made more slick and capable while still preserving battery life etc.

      Bottom line is

  • I've just started saving up for a Thinkpad T series, and was hoping that something a bit better would come out by the time I could afford it. This is great news. Hopefully it'll use the Intel 965 as I believe that this works fine with Free software drivers... I don't think the Wifi will be any good though, as usual.
    • Yes, you should be on the lookout for the upcoming ThinkPad T61, which will be Santa Rosa-based.
  • by scdeimos ( 632778 ) on Monday April 16, 2007 @07:00PM (#18758827)

    Santa Rosa carries Robson Technology, now known as Intel Turbo Memory, the flash-based disc-caching system that speeds up loading times of frequently-used data. Santa Rosa is an obvious continuation of the Centrino series.

    Maybe someone can answer this question for me: Is the flash memory for this integrated into an SMD chip on the motherboard (like the north- or south-bridge chips), or is it a plug-in module like a SIMM/DIMM?

    Flash memory wears out, the current generation only being good for a few tens-of-millions of write cycles per page. Most flash-based USB memory sticks get around this by reserving about 5% slack-space and using wear-levelling internally (similarly to JFFS). Even so, they eventually run out of usable blocks and the host computer will see block checksum errors on writing.

    If "Intel Turbo Memory" is on-chip and can't be disabled in the CMOS setup I can see people having to throw away motherboards that would otherwise be perfectly useful.

    • I can't imagine a scenario where Robson caching is unavoidable. If the flash fails, you simply no longer have use of that feature. AFAIK, the operating system of your choice has to PUT data into the cache; the system doesn't know what to put in there all by itself. With that in mind, I'd imagine that just turning it off via a control panel in Windows or a preference pane in OS X would do the trick.
    • by maxume ( 22995 )
      10 million writes on 1 gigabyte is quite a different thing than 1 million writes on 64 megabytes, and it is probably treated as read-mostly memory, so if a given block only sees a few hundred or thousand writes a day, it will last for a very, very long time.
    • Robson is a PCI Express Mini Card, but I doubt anyone will ever replace one. It's a cache; if it wears out just don't use it any more.
  • How will these new chips effect the size of new notebooks? How will they effect battery life?
    • Unless flash memory has had a number of functional improvements I'm not aware of, they will not EFFECT anything.

      But if you're asking whether they will affect size or battery life, the answer is no. Robson caching will, however, allow the hard drive greater downtime, which in turn will extend battery life, lower internal temperature (because the hard drive will not have to spin up as often), and therefore further (marginally) improve battery life in a second way thanks to lower cooling system demands and be
    • Flash memory used for disk caching will greatly extend battery life, as disk accesses that are cache hits will not require a powered down drive to spin up which uses much more power than many, many flash memory reads (not to mention is much, much slower than getting it from cache).
      So the answer is yes, they improve battery life. How much? How disk intensive is your app?
  • I can only imagine Intel is naming their products after northern CA cities. If that's the case, why the hell would they pick "Santa Rosa" as the step-up from "Napa." If anything the "Santa Rosa" platform should be a downgrade that comes with a mustache, gets your 14 year-old daughter pregnant, and emanates an overpowering manure smell when it gets hot.

    What's next? Lodi? Truckee? Daly City?
  • Why can't a multibillion-dollar company include a halfway decent graphics card in their motherboard design? Just because of that, millions of people find they can not run any games or upgrade to the latest desktop with 3D effects on their only computer.
    • by TomHandy ( 578620 ) <tomhandy@nospAM.gmail.com> on Monday April 16, 2007 @07:31PM (#18759327)
      The Intel GMA950 graphics do just fine for integrated graphics, as does GMAX3000 (part of Santa Rosa). They can both certainly play some games, and can certainly handle the 3D effects in Vista or OS X.

      I don't see why they need to have something more powerful than that though; people with more advanced needs just buy a laptop with an actual graphics card instead of integrated graphics.

      • by iamacat ( 583406 )
        Most shoppers, gift buyers and corporate purchasers do not understand what is a graphics card, RAM size or recordable DVD drive. They just expect to be sold something that will run a wide variety of applications, including the next version of the OS. Thus, the default computer should come with modern features built in as a higher priority than having the latest processor - Core Solo will do just fine instead of Core 2 Duo for most people.
  • by Michael Woodhams ( 112247 ) on Monday April 16, 2007 @07:22PM (#18759179) Journal
    There are three technologies (that I'm aware of) for using flash to cache disk. There are 'hybrid drives' where the flash is part of the hard drive, there's the Windows Vista method which uses a separately attached flash memory (typically USB), and there is this "Robson Technology" where it is on the motherboard.

    It really seems to me that the 'hybrid drive' is the Right Thing to do. The cache contents is useless without the drive, and the drive is potentially corrupt without the cache contents, so why make them separable? With appropriate firmware, the hybrid drive can make the existence of the cache transparent to the OS, so no OS support is required (but you can allow the OS finer control over the cache if it does support it.) You also automatically add more cache as you add more drives.

    (Incidentally, I hope MS doesn't have a patent on this - I thought of it years ago, and I'm not even an engineer.)

    I can see the Windows method as a useful 'stop-gap' to get the benefit with a non-hybrid drive, but if you're buying new hardware anyway, why would you want to put the cache on the motherboard instead of the hard drive? The only advantage I can think of is that if you have multiple drives, you can dynamically allocate how much cache is associated with each drive, according to usage patterns.
    • There are 'hybrid drives' where the flash is part of the hard drive, there's the Windows Vista method which uses a separately attached flash memory (typically USB), and there is this "Robson Technology" where it is on the motherboard.

      The hardware and OS support is orthogonal; Vista supports all three kinds of flash.

      It really seems to me that the 'hybrid drive' is the Right Thing to do. The cache contents is useless without the drive, and the drive is potentially corrupt without the cache contents, so why ma
      • If the cache is write-through then you wouldn't have to worry about drives getting corrupted.

        Fair point.

        You can remove a hard disk from a notebook, but it's not exactly easy so I don't think that's much of a concern.

        Power saving is just one reason for flash cache, so it isn't (or shouldn't be) restricted to notebooks. On desktop, the major advantage would be boot speed, with general speed and noise reduction as secondary benefits. I expect you'd use more of it for reading and less for writing than for a la
    • by the7cs ( 667215 )
      Where I hope to see this go is larger flash memory on the mother board, to the end that your corporate base OS image, and all of your application source installation files are stored on it. I would also see it going to the point where you can perform a data backup to the flash memory.

      If you lost a hard drive or had to reimage this would be very useful for a mobile workforce.
  • Linux? (Score:3, Insightful)

    by DoofusOfDeath ( 636671 ) on Monday April 16, 2007 @08:29PM (#18760223)
    Any news on how hard Intel will work to ensure that good free software exists for driving Santa Rosa's wifi, wired ethernet, and video chips?
  • by vocaro ( 569257 ) *
    #21 on their list should have been the block-proof DHTML pop-up ad on the second page of their article.
    • Or (Score:2, Funny)

      Or, quite possibly, discovering that you replied to the wrong article the very moment you hit "Submit"...
  • by TheDarkener ( 198348 ) on Monday April 16, 2007 @09:02PM (#18760761) Homepage
    I can say that this lineup will come, by default, with wine glass, extra-large chrome SUV rims, and Starbucks coffee holder.

    Oh, and it'll still whine about how little it has.
  • Turbo memory ... WTF ?? I worked at this pop-stand (Intel) in R+D and and watched this technology (code named Boxcar at the time) be developed starting back around 2003 with the original idea being just what you see here - a local cache for the hard drive. However I have to say that Turbo memory is about the lamest freakin name ... marketing at Intel always was a total crap shoot now it appears to be a total crapper. And I have no idea who this Robinson character is or if this Turbo crap is his doing.

    Many
  • So when will these new processors start showing up on the Dell and HP websites?

Our business is run on trust. We trust you will pay in advance.

Working...