Results 1 to 20 of 20

Thread: Nehalem Information

  1. #1
    Xtreme Addict
    Join Date
    Jul 2006
    Location
    Vancouver, BC
    Posts
    2,061

    Nehalem Information

    In case you are intersted, I'm compiling a list of information related to Nehalem on my blog...

    www.nehalemnews.com

    I'm sure most XS members interested in Nehalem are also aware of these recent threads... (from most recent)

    Nehalem Clock and Power Domains

    Nehalem EP - Bloomfield Super Pi Benchmark

    Bloomfield Specs

    Nehalem Motherboards

    Nehalem Clocks and "Turbo Mode"

    Article: Nehalem 101

    Rumor: No Overclocking on Mainstream Nehalem

    Sighting: Nehalem working at 3.2GHz

    Sighting: Intel Bloomfield LGA1366 Motherboard

    I'll add others as they appear.
    Last edited by virtualrain; 06-17-2008 at 12:56 PM.

  2. #2
    Xtreme Member
    Join Date
    Oct 2007
    Posts
    355
    I'm interested, thanks for pooling all that into one source... Don't like that overclocking bit!
    CPU: Lapped L727A861 Q6600 3.61GHz (450x8) (1.275VID) @ 1.700V BIOS. WC w/Apogee GTX + Cu top + custom backing plate
    GPU: EVGA 8800GT. WC w/MCW60 + RAMsinks
    MB: ASUS Maximus Formula: Rampage Formula BIOS 0402. NB @1.61V BIOS. WC w/MCW30. FSBtV=1.38V BIOS. CPUPLLV=1.82V BIOS
    RAM: Corsair Dominator PC2 8500 4x1GB 1066@1081. 2.14V BIOS

  3. #3
    Xtreme Addict
    Join Date
    Jun 2004
    Location
    near Boston, MA, USA
    Posts
    1,955
    Thanks for compiling the info. While I don't think it will end up being true, it is sad to even contemplate that Intel might once again go back to limiting overclocking on anything but the "extremes". Like similar rumours in past generations it probably won't pan out, but there is always a worry they might abuse their upper hand in the market.

    Thank you kindly though, for putting information I am very much looking forward to, in one place.

  4. #4
    Xtreme Member
    Join Date
    Mar 2008
    Location
    Germany
    Posts
    351
    Stickify it!

    EDIT: You've been star'ed
    Last edited by Neuuubeh; 04-28-2008 at 01:24 AM.
    X3350 | DFI LP X38 T2R | d9gkx
    9800gtx | Raptor1500AHFD/5000AACS/WD3201ABYS
    Corsair 620HX | Coolermaster CM690

  5. #5
    Xtreme Addict
    Join Date
    Mar 2007
    Posts
    1,377
    This is interesting. Moreso than intel locking it out, I'm curious how the shrinking architecture will stand up to current overclocking methods. As it gets smaller it gets more fragile to voltage regardless of heat.

  6. #6
    Registered User
    Join Date
    Mar 2008
    Location
    Wherever I may roam...
    Posts
    591
    I almost went Intel on my current build (and I am considering Intel for next build) but not if they limit OCing

  7. #7
    Xtreme Addict
    Join Date
    Jul 2006
    Location
    Vancouver, BC
    Posts
    2,061
    EDIT: See next post
    Last edited by virtualrain; 04-29-2008 at 11:07 AM.

  8. #8
    Xtreme Addict
    Join Date
    Jul 2006
    Location
    Vancouver, BC
    Posts
    2,061
    I recently blogged the following article and I'm interested in your opinion...

    A big question on everyone's mind is what CPU frequency will intial Nehalem parts be clocked at and what overclocking head room might exist?

    Although a demo machine was recently reported to be running Nehalem at 3.2GHz, we can't be positive based on the evidence provided that this early sample was, in fact, running at that speed. However, it's not unreasonable for Bloomfield to launch at 3GHz speeds given that Penryn's highest binned parts are shipping at 3.2GHz. This is supported, in part, by the fact that Nehalem's 731 million transistors compares favorably to Penryn's 820 million (both at 45nm).

    It seems plausible that Nehalem should theoretically clock just as well as Penryn at the same Thermal Design Power (TDP) with the same cooling solution. The big unknown is what effect the onboard Memory Controller Hub (MCH) will have on clock speed limitations.

    Another intersting aspect to Nehalem is the reports of a "Turbo Mode". While published details are hard to come by, this dynamic core clocking capability is illustrated in the slide below (courtesy of HKEPC). It appears to be an extension of the Performance States of the Adavanced Configuration and Power Interface (ACPI) specification also know as SpeedStep technology (P-States). It suggests that when load allows one or more cores to be throttled down into a low frequency mode (LFM) the remaining active loaded cores can actually be overclocked to higher than default clock frequencies as long as the default TDP is not exceeded.


    (source: HKEPC)

    While this is an exciting development, unfortunately it raises more questions than it answers.... for example, it's not clear if this feature will work with overclocked chips or only those running at default clocks and multipliers. It's also not clear what events or conditions trigger the LFM for the unutilized cores and similarly what events or conditions trigger the overclocked P-states of the active cores. Finally, there's no insight into how much control the BIOS, OS, or end-user will have over this capability.

    We can only hope that Nehalem's Turbo Mode follows the current SpeedStep implementation which can be managed at both the BIOS and/or Operating System level to allow modifying the multiplier in response to CPU loads. That is, if one overclocks their reference clock from default, this feature will ideally manage core clocks by adjusting the multi's up or down as load dictates thus allowing overclockers to benefit from this feature as much as factory clocked systems.

    If this feature does support overclocking, it will add another dimension to stability testing as a stable operating point for all 4 cores will also have to consider the "Turbo Mode" state.

    Needless to say, the benefits of this kind of dynamic overclocking with balanced performance on multiple cores for multi-threaded apps while also offering maximum clock speed on single-threaded programs, all with the same cooling solution, effectively ensures one can "have their cake and eat it too"!
    Thoughts?

  9. #9
    Xtreme Member
    Join Date
    Oct 2007
    Posts
    355
    Well, OC'rs will always pay way more stricter attention to detail when building a rig as apposed to a "mass produced" machine. So, to put us OC'rs out of the equation by saying "...this feature will ideally manage core clocks by adjusting the multi's up or down as load dictates thus allowing overclockers to benefit from this feature as much as factory clocked systems.", really sucks. But then maybe it is a limitation of the thinner 45nm. A certain width of conductor can only carry so much heat... Which brings me to a point. IF they would throttle from 2 to 4 cores, with the advantage being at 2 cores=more speed. I wouldn't hesitate from buying this chip. Only prob is, can they separate the 2 cores on the 2 dies? spreading the heat better. I know a quad is made of 2 separate dual cores; what I am saying is if they can shut down 1/2 of each dual core and activate 1 cpu from each core, then the heat would be dissipated much better, i.e., they would be 2x further apart in dual mode.
    Last edited by ElEctric_EyE; 04-29-2008 at 03:48 PM. Reason: misspellings...
    CPU: Lapped L727A861 Q6600 3.61GHz (450x8) (1.275VID) @ 1.700V BIOS. WC w/Apogee GTX + Cu top + custom backing plate
    GPU: EVGA 8800GT. WC w/MCW60 + RAMsinks
    MB: ASUS Maximus Formula: Rampage Formula BIOS 0402. NB @1.61V BIOS. WC w/MCW30. FSBtV=1.38V BIOS. CPUPLLV=1.82V BIOS
    RAM: Corsair Dominator PC2 8500 4x1GB 1066@1081. 2.14V BIOS

  10. #10
    Xtreme Addict
    Join Date
    Jul 2006
    Location
    Vancouver, BC
    Posts
    2,061
    Quote Originally Posted by ElEctric_EyE View Post
    Well, OC'rs will always pay way more stricter attention to detail when building a rig as apposed to a "mass produced" machine. So, to put us OC'rs out of the equation by saying "...this feature will ideally manage core clocks by adjusting the multi's up or down as load dictates thus allowing overclockers to benefit from this feature as much as factory clocked systems.", really sucks. But then maybe it is a limitation of the thinner 45nm. A certain width of conductor can only carry so much heat... Which brings me to a point. IF they would throttle from 2 to 4 cores, with the advantage being at 2 cores=more speed. I wouldn't hesitate from buying this chip. Only prob is, can they separate the 2 cores on the 2 dies? spreading the heat better. I know a quad is made of 2 separate dual cores; what I am saying is if they can shut down 1/2 of each dual core and activate 1 cpu from each core, then the heat would be dissipated much better, i.e., they would be 2x further apart in dual mode.
    Nehalem will be a single piece of silicon with 4 cores on it (Nehalem FAQ). Also, it doesn't matter what particular cores are shutdown or active or where they are physically located in the package for this TDP budgeting to work effectively.

    The Turbo Mode basically works like this. If 4 cores running at default clocks generates 4X in heat, then if you shut down 2 cores you would generate something like 3X in heat (as there are things like cache and IMC still drawing power). This gives you 1X in heat headroom to overclock your 2 active cores so that in the end you are still producing only 4X in heat but with 2 overclocked cores instead of 4 cores running at stock speeds.

    Now if your cooling is a lot better than stock, you can do all of this with an overclocked processor.

    For example, lets say your new Nehalem runs stock at 200x16 or 3.2GHz (remember there's no FSB just a referenece clock and I'm assuming for example that it's 200MHz)... and you then overclock it to a modest 3.6GHz (225x16) because you have good aftermarket cooling that can handle a higher than stock TDP. So you enable Turbo Mode and you find that when running single threaded apps, 2 cores are shutdown and the other two run at 225x18 or 4GHz... all without temps increasing! Nice! What overclocker wouldn't want such a feature?
    Last edited by virtualrain; 04-29-2008 at 11:36 PM.

  11. #11
    Xtreme Member
    Join Date
    Oct 2007
    Posts
    355
    Ok, now I understand abit better, so basically it sounds like we are to get the best of both worlds from the core 2 duos and core 2 quads + the on die memory controller. Will some boards take off from where Skulltrail left off at an 8 core mobo?
    CPU: Lapped L727A861 Q6600 3.61GHz (450x8) (1.275VID) @ 1.700V BIOS. WC w/Apogee GTX + Cu top + custom backing plate
    GPU: EVGA 8800GT. WC w/MCW60 + RAMsinks
    MB: ASUS Maximus Formula: Rampage Formula BIOS 0402. NB @1.61V BIOS. WC w/MCW30. FSBtV=1.38V BIOS. CPUPLLV=1.82V BIOS
    RAM: Corsair Dominator PC2 8500 4x1GB 1066@1081. 2.14V BIOS

  12. #12
    Xtreme Addict
    Join Date
    Jul 2006
    Location
    Vancouver, BC
    Posts
    2,061
    Quote Originally Posted by ElEctric_EyE View Post
    Ok, now I understand abit better, so basically it sounds like we are to get the best of both worlds from the core 2 duos and core 2 quads + the on die memory controller. Will some boards take off from where Skulltrail left off at an 8 core mobo?
    It's not clear if anyone will make a dual socket enthusiast board. There will certainly be dual socket server boards. Later there will also be 8-core Nehalems so imagine a dual socket board loaded with 16 cores and 32 threads of goodness! Such a system might even play Crysis on high settings!

  13. #13
    Xtreme Addict
    Join Date
    Jul 2006
    Location
    Vancouver, BC
    Posts
    2,061
    Just wanted to let you know I posted a new article on how the FSB is limiting memory performance and the benefits an integrated memory controller will bring to memory performance...

    As you can see, a FSB frequency of 333MHz currently available on Intel's top of the line products cannot even support the full bandwidth capabilities of widely used interleaved dual-channel DDR2 memory... The FSB is already a bottleneck when running DDR2-800 memory in dual channel mode.

    The prospects for the FSB are even more dire when you look at DDR3 performance. The current 333MHz FSB is only suitable for single-channel DDR3 up to DDR3-1333 speeds. Running multi-channel or higher speed DDR3 memory causes the current FSB to become a serious bottleneck.
    Full Story

  14. #14
    Xtreme Addict
    Join Date
    Jul 2006
    Location
    Vancouver, BC
    Posts
    2,061
    I added a new post today on the likely clock and multi implementation for Nehalem...

    Hopefully this article provides some insights into what to expect with regards to how clocking and overclocking will be implemented on Nehalem. In general, overclocking should be much simpler in terms of controlling the individual clock domains. What's not clear is how much end-user control Intel will provide over these components and what intricies or trade-offs may be introduced as a result of the on-die IMC. It's also unclear how much overclocking head-room we can expect from Nehalem compared to the riches we've enjoyed with the Core2 generation. The far more complex design of Nehalem and sheer die-size may pose a number of challenges to overclockers, the least of which is added stress on the cooling system. However, Intel's superior transistor technology and Hi-K metal gate 45nm process should allow for much more clocking head-room than we've witnessed with AMD's 65nm native quad.
    Full Story

  15. #15
    Xtreme Cruncher
    Join Date
    Feb 2008
    Location
    C:\WINDOWS\system32\
    Posts
    1,451
    OK, subscribed. I'm curious.

  16. #16
    Registered User
    Join Date
    May 2005
    Location
    UK - by the sea
    Posts
    42
    Quote Originally Posted by virtualrain View Post
    I recently the following article and I'm interested in your opinion...

    "Details on Nehalem's Turbo Mode"
    Thanks for the post, it’s the first time I’ve seen this info. It’s an extension of what Intel already offers on their mobile platform. I had a Centrino laptop with a 65nm 2GHz 800MHz FSB CPU and that supported running at the next highest multiplier if only 1 core was loaded so it could max at 2.2GHz. This looks a lot more flexible and promising.

  17. #17
    Xtreme Member
    Join Date
    Jul 2007
    Location
    now
    Posts
    242
    are there going to be two different sockets for desktops: LGA1366, LGA1160 & consequently, no upgrade path between the two??
    ------media machine------
    Q6600 L737B || EVGA 780i || 8GB G.Skill DDR2-1000 PQ || 2 x EVGA 9800 GX2
    150GB Raptor (Vista x64 SP1) || 3 x 640GB WD6400AAKS RAID0

    LG Blu-ray/HD-DVD Combo || Dell 3008WFP || Corsair HX1000W || Rocketfish
    Air Cooled with ThermalRight Ultra 120 Extreme

  18. #18
    Xtreme Addict
    Join Date
    Jul 2006
    Location
    Vancouver, BC
    Posts
    2,061
    Quote Originally Posted by lefy View Post
    are there going to be two different sockets for desktops: LGA1366, LGA1160 & consequently, no upgrade path between the two??
    That's correct, the performance models (Bloomfield) support tri-channel DDR3 which require the extra pins. The mainstream models (Lynnfield/Havendale) will only offer dual-channel DDR3 so can work in a smaller, cheaper, less complex socket.

  19. #19
    Xtreme Addict
    Join Date
    Jul 2006
    Location
    Vancouver, BC
    Posts
    2,061
    Added some new threads to the OP for interested readers to check out. Lots of new stuff at
    NehalemNews
    recently as well.

    Cheers,
    -Chris.

  20. #20
    Registered User
    Join Date
    Jun 2008
    Posts
    20
    Allright I am looking to make a new system in Augest, and I was wondering would it be worth waiting for a Nehelm system or just get a dual e5420 system? I'd like to have at least 8 cores, with head room to upgrade to better CPUs way down the line.

Bookmarks

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •