Future MacBook Pro-bound Intel chips due next spring

Posted:
in Future Apple Hardware edited January 2014
The next time Apple will have the opportunity to boost the processor specifications of its MacBook Pro line will be next spring, when Intel pushes out a final update to its Montevina platform consisting of two high-performance mobile chips, according to reports.



Apple's current MacBook Pros employ a "P" series, medium voltage (25W) P8600 2.4GHz chip at the low end, and a "T" series, standard voltage (35W) T9400 2.53GHz chip at the high end.



The T series chip costs the company roughly $32 less than the P series, according to pricing estimates published on the web, but runs slightly hotter. Apple also offers a build-to-order option on the MacBook Pro that lets customers choose a "T" series, standard voltage (35W) T9600 2.8GHz chip for a $300 premium. The chip costs the company roughly $200 to $250 more than the 2.53GHz variants, according to estimates.



Citing sources at Taiwanese notebook makers, DigiTimes claims Intel will refresh these Montevina-based Core 2 Duo offerings in April by adding two new variants: a P8800 running at 2.66GHz and a T9900 clocking in at 3.06GHz.







VR-Zone earlier in the week provided corroborating evidence to this end via a series of leaked Intel roadmap presentation slides. The visuals list the two chips as arriving anytime from mid-to-late second quarter, which runs April through June.



At this juncture, Apple could presumably bump its 15-inch MacBook Pro configurations up to standard frequencies as high as 2.66GHz and 2.8GHz, with a 3.06GHz build-to-order option. Alternatively, it could inch standard configurations up to 2.53GHz and 2.66GHz, leaving the 2.8GHz and 3.06GHz chips as potential build-to-order choices. It should be noted, however, the 17-inch unibody MacBook Pro should be in the mix by this time, potentially adopting one of the remaining high-end chips, such as the 2.8GHz.



The current array of Intel Core 2 Duo mobile chips employed in Apple's unibody MacBook Pros.



The spring refresh is expected to be the last to hit Intel's Montevina platform before the chipmaker turns to its new and broad Nehalem architecture, which has since officially been dubbed the "Core i7." Calpella is the code-name for Intel's Core i7-based mobile platform, which will umbrella chips currently referenced by the "Clarksfield" code-name (the same way that Montevina umbrellas mobile chips from the Penryn family). [Dizzy yet?]



Spring 2009 Core 2 Duo mobile chip additions likely to be adopted by Apple's MacBook Pros.



With the launch of Calpella not scheduled until the third quarter of 2009 at the earliest, few if any hard specifications are known about its new mobile chips. More generally, however, it's been speculated that Calpella processors will be based on a 32-nanometer manufacturing process and possibly abandon the use of separate northbridge and southbridge chipsets for components that will be integrated into some of its processor families.



A bit earlier next year, Intel also plans to introduce its new GM47 integrated graphics chipset with a 1066MHz front-side bus and 640MHz graphics core frequency. Given Apple's emphatic endorsement of NVIDIA's new chipsets across its notebook lines, it's unlikely the company holds any plans to adopt the new Intel part.



Update: Charts and story text updated to reflect confirmation that the 2.53GHz Intel processor employed by the unibody MacBook Pro is the "T" series, standard voltage (35W) T9400, and that Apple is not using the "P" series, medium voltage (25W) P9500.
«13

Comments

  • Reply 1 of 46
    I feel like such a nerd because I actually understood that article. Ugh.
  • Reply 2 of 46
    hmurchisonhmurchison Posts: 12,424member
    What a colossal embarassment for Intel to have nVidia

    come and eat their lunch like they have. I guess it's good

    comeuppance for Intel to get their noses rubbed in it since

    they haven't had to worry about AMD doing it in recent years.



    Core i7 (just rolls off the tongue doesn't it?) mobile coming

    late 09 isn't a bad thing. Montevina isn't a bad and Apple needs

    to have Snow Leopard up to snuff once Nehalem hits server, desktop

    notebooks.



    I'm curious to know what nVidia does for an MCP79 encore. I don't

    know how long they keep their chipsets in the market but I figure that

    by mid 2010 they'll have a successor. Will it be enough to keep Apple on

    their chipset? We'll see.
  • Reply 3 of 46
    lorrelorre Posts: 396member
    Is it just me or are those images really distorted and unreadable? Also the main site logo is farked up...
  • Reply 4 of 46
    Quote:
    Originally Posted by tecknojoe View Post


    I feel like such a nerd because I actually understood that article. Ugh.







    I was about to say... this is all such "inside baseball" stuff. Consumers don't care one bit about it.
  • Reply 5 of 46
    kasperkasper Posts: 941member, administrator
    Quote:
    Originally Posted by SpamSandwich View Post






    I was about to say... this is all such "inside baseball" stuff. Consumers don't care one bit about it.



    We're called Apple"inside"r =P



    K
  • Reply 6 of 46
    paxmanpaxman Posts: 4,729member
    Quote:
    Originally Posted by tecknojoe View Post


    I feel like such a nerd because I actually understood that article. Ugh .



    .
  • Reply 7 of 46
    HAHA, ok ok. back on track. So with everything being said about the macbook pros right now, does anyone think that macworld is going to show any new addition/upgrade to the 15" MBP(besides what just happened to them)? Or is it likely that they will not see any more attention until the new chips?
  • Reply 8 of 46
    Hopefully Apple will move to standardize on 25W TDP parts in the next refresh with 25W 2.53GHz and 2.66GHz processors on the 15". The 17" can use the 35W 2.8GHz model, while the 3.06GHz 35W part can be a BTO for both 15" and 17" MBP.



    From what I've heard, Apple appears to be using 35W parts for there current 2.53GHz models which is disappointing. May well be why some are seeing Apple's 5hr battery life claims, namely the low-end 15" MBP, while others aren't seeing it, namely the high-end 15" MBP.



    And in regards to Capella, it'd be interesting to see if it'll end up being based on 32nm Westmere rather than 45nm Nehalem. The first 32nm parts are supposed to show up in Q4 2009, so a 1 quarter delay to move Capella to 32nm may well be worthwhile. AMD is offering no competitive pressure in notebooks, especially now that nVidia IGPs have negated the AMD's Puma platform IGP advantage. And Nehalem's design features like integrated memory controllers have no real benefit for notebooks. Nehalem itself is design primarily to address Intel's weakness in the server market where having memory channels and memory bandwidth scaling with the number of processors is important. Being on the same 45nm process means Nehalem won't bring much power consumption improvement to notebooks and won't be able to enable mainstream mobile quad cores seeing current Penryn based mobile quad cores require an Extreme Edition 45W TDP rather than the standard 35W or 25W TDPs of mobile dual cores. Launching using a 32nm process for mobile chips makes sense to give strong power savings and reduce cost to allow mainstream mobile quad cores.
  • Reply 9 of 46
    Intel has been talking about their "tic toc" method of chip delivery for a few years now. The "tic" is the die shrink delivery and "toc" is the core delivery.



    The Penryn was the "tic" part of the current delivery schedule (existing core just a die shrink), and Nehalem will be the "toc" portion where the 45nm process is used on the new core microarchitecture.



    Intel uses the "tic toc" process to help ensure reliable products. Intel had reliability problems many years ago (and more than once I might add) that occurred when they would deliver a new core on a new die shrink process. After a few painful, and public, issues Intel decided to change to the "tic toc" method to reduce changes for each cycle, this reducing the chance of delivering a bad product.



    The first Nehalem architecture ships arrive next month for servers. Desktops arrive in mid-2009 and Laptops about one quarter later.



    For more info on Nehalem go to: http://www.intel.com/technology/arch..._brand|kEB7D|s



    The article specifically states that Nehalem will be delivered using the 45nm process.



    The 32nm die shrink is scheduled to arrive sometime in the second half of 2009 and should start with the lower-volume products like servers or desktops (typical of Intel). The Nehalem architecture will of course be used in the 32nm process. And the cycle continues...............
  • Reply 10 of 46
    pjapja Posts: 6member
    Quote:
    Originally Posted by lecube View Post


    Desktops arrive in mid-2009 and Laptops about one quarter later.



    Seriously??



    I need a new Mac Pro ASAP, but having waited a year to replace the PowerMac was hoping they'd launch a new one before or after Xmas with the Nehalem architecture which I believe would necessitate a redesign case (hopefully a little smaller)
  • Reply 11 of 46
    Quote:
    Originally Posted by tecknojoe View Post


    HAHA, ok ok. back on track. So with everything being said about the macbook pros right now, does anyone think that macworld is going to show any new addition/upgrade to the 15" MBP(besides what just happened to them)? Or is it likely that they will not see any more attention until the new chips?



    I expect Macworld will be the debut of the 17" unibody MacBook Pro and the introduction of the Nehalem platform in the Mac Pro and XServe. New software will round out the show.
  • Reply 12 of 46
    satchmosatchmo Posts: 2,699member
    So I guess spring is when we'll see the optional matte screen version then.
  • Reply 13 of 46
    The tables say that the T9600 costs ~$200 more PER ONE THOUSAND CHIPS. That is far from a $200 increase per single chip.



    Edit: Increased cost compared to P9500, T9400
  • Reply 14 of 46
    solipsismsolipsism Posts: 25,726member
    According to CPU-Z in Windows on a new MBP, it is the 35W T9400.
  • Reply 15 of 46
    hmurchisonhmurchison Posts: 12,424member
    Quote:
    Originally Posted by pja View Post


    Seriously??



    I need a new Mac Pro ASAP, but having waited a year to replace the PowerMac was hoping they'd launch a new one before or after Xmas with the Nehalem architecture which I believe would necessitate a redesign case (hopefully a little smaller)



    No Core i7 should be launching this year with the mobile parts rumored to be delayed until late 09.



    http://www.maximumpc.com/article/new..._rumor_week_46
  • Reply 16 of 46
    mr. hmr. h Posts: 4,870member
    Quote:
    Originally Posted by honkytonk View Post


    The tables say that the T9600 costs ~$200 more PER ONE THOUSAND CHIPS. That is far from a $200 increase per single chip.



    Edit: Increased cost compared to P9500, T9400



    You've misunderstood.



    You don't think people get 1000 T9600s for $530? That'd be only 53 cents each! What the datum is telling you is that if you buy 1000, they'll cost $530 each. If you buy fewer than 1000, each one will cost you more. If you buy a lot more than 1000 (e.g. 1,000,000), the price per processor comes down, possibly quite substantially (say $450 each as a wild guess, Intel only publish "per 1000" prices).
  • Reply 17 of 46
    mr. hmr. h Posts: 4,870member
    Quote:
    Originally Posted by appleinsider View Post


    confirmation that the 2.53ghz intel processor employed by the unibody macbook pro is the "t" series



    Lame!
  • Reply 18 of 46
    Quote:
    Originally Posted by lecube View Post


    Intel has been talking about their "tic toc" method of chip delivery for a few years now. The "tic" is the die shrink delivery and "toc" is the core delivery.



    The Penryn was the "tic" part of the current delivery schedule (existing core just a die shrink), and Nehalem will be the "toc" portion where the 45nm process is used on the new core microarchitecture.



    Intel uses the "tic toc" process to help ensure reliable products. Intel had reliability problems many years ago (and more than once I might add) that occurred when they would deliver a new core on a new die shrink process. After a few painful, and public, issues Intel decided to change to the "tic toc" method to reduce changes for each cycle, this reducing the chance of delivering a bad product.



    The first Nehalem architecture ships arrive next month for servers. Desktops arrive in mid-2009 and Laptops about one quarter later.



    For more info on Nehalem go to: http://www.intel.com/technology/arch..._brand|kEB7D|s



    The article specifically states that Nehalem will be delivered using the 45nm process.



    The 32nm die shrink is scheduled to arrive sometime in the second half of 2009 and should start with the lower-volume products like servers or desktops (typical of Intel). The Nehalem architecture will of course be used in the 32nm process. And the cycle continues...............



    In the tic-toc model does not necessitate that the new chip be introduced in server or high-end desktop. Intel has only started doing that because servers are where AMD is still competitive so Intel has decided to put pressure there first.



    The tic-toc model actually began with on notebooks with Yonah (Core Duo in marketing speak) which launched Intel's 65nm process and was used in the the first Intel Macs. What I'm saying is that since mobile Nehalem is coming so close to the arrival of the 32nm shrink, it is worthwhile to just cancel the mobile Nehalem and launch 32nm Westmere in the mobile market as was done with Yonah. It may not be glamourous, but Penryn can easily last until Westmere seeing that there is no pressure from AMD, and mobile Nehalem itself isn't particularly beneficial to notebooks since it won't bring much power improvements or encourage the move to mobile quad cores since quad core Nehalem costs more than quad core Penryn since Nehalem is a single, big die, while quad core Penryns use 2 smaller dies. Westmere has doesn't have these disadvantages and coming so close to mobile Nehalem, makes mobile Nehalem all the more disappointing if we have to wait nearly a year until Q3 2009 for mobile Nehalem which will then be a generation behind when Westmere arrives a quarter later.
  • Reply 19 of 46
    justflybobjustflybob Posts: 1,337member
    Quote:
    Originally Posted by satchmo View Post


    So I guess spring is when we'll see the optional matte screen version then.



    I know there has been much anguish and wringing of hands over this, but after extensive testing at the local Apple Store I found the reflection issues to be way overblown. Mind you, I did the same comparison when I bought my current 2.33 MBP 17 and at that time the difference WAS major.
  • Reply 20 of 46
    Quote:
    Originally Posted by AppleInsider View Post


    Spring 2009 Core 2 Duo mobile chip additions likely to be adopted by Apple's MacBook Pros.



    Based on your chart, AppleInsider seems like it's hoping for the 2.66GHz P8800 to be used in MacBook Pros. In truth though the 2.66GHz chip that we should be supporting is the 2.66GHz P9600 coming in Q1 2009. The 9xxx series all have 6MB of L2 cache while the 8xxx series all have 3MB L2 cache. Intel always does this where they introduce a fully equipped processor at a certain clock speed and come up with a cut down one a few months later at the same clock speed hoping no-one will notice. Just like the T9550 that is on the roadmap is also a 2.66GHz part, but most likely has VT support removed to justify the 9x50 designation. Most consumers don't since despite focus on multi-core and architectural improvements, clock speed still sells.



    I really wish Apple went with 6MB L2 cache processors across the board on the MacBook Pro. If they wanted something to distinguish between the premium MacBook Pro line and the MacBook line, especially now that they are so closely equipped, L2 cache seems like a good way to do it. Granted Intel doesn't offer a 2.4GHz model with 6MB L2 cache, but really, if Apple wanted it they could get it. Afterall just look at the current iMacs. They are supposed to be a special run of Penryns, but they aren't simply Penryns on an overclocked FSB. If you look at their L2 cache, they are all 6MB versions. Even the 2.4GHz version on the iMac has a 6MB L2 cache, yet no Intel mobile Penryn at 2.4GHz offers 6MB L2 cache, only 3MB L2 cache. So Intel specifically made a 6MB L2 cache 2.4GHz processor for Apple.



    Hopefully Apple's next notebook refresh uses 25W TDP, 6MB L2 cache processors standard in the 15" MBP, which would be the 2.53GHz P9500 and the 2.66GHz P9600. Not any 3MB 8xxx variants.
Sign In or Register to comment.