2.66 Quad vs 2.93 Quad Buyer's Guide (Includes Octos)

Discussion in 'Mac Pro' started by tome viewer, Apr 25, 2009.

  1. macrumors newbie

    Joined:
    Feb 27, 2009
    #1
    2.66 Quad Core vs 2.93 Quad Core

    Bragging rights only, does not show “real world” performance difference (Courtesy OWC):

    [​IMG]

    CINEBENCH
    14251
    15800
    10.86941267279489%

    GEEKBENCH
    9390
    10236
    9.00958466453674%


    Real World Single Thread Difference:

    [​IMG]

    3x2GB Sticks – Total Ram 6GB
    28.78
    27.28
    5.211952744961779%

    4x2GB Sticks – Total Ram 8GB
    28.82
    27.66
    4.02498265093685%



    Real World Multi Thread Difference:

    [​IMG]

    3x2GB Sticks – Total Ram 6GB
    177
    165
    6.779661016949152%

    4x2GB Sticks – Total Ram 8GB
    177
    163
    7.909604519774012%



    Recent BareFeats Benchmarks:

    GRAPH LEGEND
    m2.9 4c = 4-core 2009 Mac Pro 2.93GHz with 12G of RAM and Radeon HD 4870
    m2.7 4c = 4-core 2009 Mac Pro 2.66GHz with 6G of RAM and Geforce 120 GT

    [​IMG]

    7.344632768361582%

    This percentage gain is misleading because if the 2.66 were equally equiped with 12GB Ram the percentage gain of the 2.93 over the 2.66 would now only be 6.395348837209303% in this multi threaded test.


    [​IMG]

    0% Difference

    [​IMG]

    0% Difference



    Our Testing:

    In real world use it is impossible to detect a performance difference between the 2.66 and 2.93. And I need to emphasize the word “impossible”. Five of us used both configurations for an hour each and not a single one of us could tell the difference. Your money is better spent elsewhere. What can you buy with the extra $500? Three 1TB WD Blacks (you can use the stock 640GB Blue for TM) plus add 6GB Ram and have $75 in extra change. $500 off a ACD. A SSD for $300-360 and the 4870 upgrade with it. Once we jumped on a 2.66 with SSD we could feel a dramatic performance difference. The difference is very significant and not minor like the processors. In regard to the VelociRaptor's I would stay away from them for the same insignificant speed bump difference as the processors. Here is a quote from a Raptor user:

    “Not a quantum leap over a 32M cache 7200k

    Pros: The fastest consumer desktop SATA hard drive money can buy. I was expecting a bit more speed, however, all in all this is a pretty good drive. Runs very cool, cooler than my 750GB caviar black I use now as a storage drive since I got the raptor. I also think its pretty quiet for the most part as I cannot hear it over my case fans, which are of medium loudness.

    Cons: To me 230$ is a bit hard to swallow for the incremental performance over a fast, modern 7200K 32M drive. I gave this 4 starts simply because of its price/performance. Here are some quick results using HD Tune 2.55 vs. my WD 750GB WD7501AALS caviar black.

    Average Transfer rate:
    VelociRaptor - 101.4 MB/s
    750GB Caviar Black - 90.5 MB/s

    Burst Rate:
    VelociRaptor - 182.2 MB/s
    750GB Caviar Black - 143.1 MB/s

    Access Time:
    VelociRaptor - 7.0ms
    750GB Caviar Black - 11.1ms

    Modern 32MB cached 7200k drives like a WD Caviar Black series are not too far behind the velociraptor in benchmarked speed and you will be hard pressed to notice a real world difference.

    Other Thoughts: Price is kinda steep for marginal speed increase over a 32MB cached caviar black. Makes you want to think about shelling out some more $$ and just take the SSD plunge. A newer SSD like the new OCZ Vertex or even Intel's SSDs wipe the floor with a conventional HD like a VelociRaptor. In my humble opinion, unless you have money to burn and don't want to get an SSD, pass on a velociraptor and get a large capacity 32MB cached Caviar Black as you will notice zero real world performance difference, save a lot of money and have much greater capacity. Again the VelociRaptor is a great fast SATA drive, but after my experience, when the rubber meets the road, this drive is not at all worth the price premium over a fast 7200k drive. If extreme storage speed is what your after, especially a drive that will serve as a boot drive or gaming drive, nothing short of a new high performance SSD will satisfy you. “​

    Note: To make matters worse, the 750GB Black that he used in his test is the worst performer of the blacks. The 500GB and 640GB are faster, and the 1TB the fastest. In other words, the VelociRaptor is even less appealing when pitted against one of the "better" WD Blacks.

    Remember all of this is just our opinion and we would love to hear feedback from other users.

    :)
     
  2. macrumors 6502

    Joined:
    May 24, 2007
    #2
  3. macrumors member

    Joined:
    Jul 17, 2002
    #3
    Why do the After Effects Render Tests show negative percentage results? Is finishing the test faster supposed to be a bad thing in their eyes? Same with the After Effects CS4 benchmarks from Barefeets...they give that a negative percentage result and yet the 2.93 QC finished 13 seconds faster.

    best,
    Jon
     
  4. macrumors 601

    Tesselator

    Joined:
    Jan 9, 2008
    Location:
    Japan
    #4
    Almost all the information including even the name of the thread is very confusing. Nothing is easily discernible or straight forward. You get an A for effort and an F- for presentation. :)
     
  5. macrumors 65816

    zer0tails

    Joined:
    Mar 23, 2008
    Location:
    Canada
  6. macrumors 603

    VirtualRain

    Joined:
    Aug 1, 2008
    Location:
    Vancouver, BC
    #6
    LOL... this first test proves that a 10% clock increase provides approximately a 10% increase in benchmark scores... you can call that bragging rights, proof that the benchmarks work, or whatever you want... but it is an increase in performance directly proportional to the increase in clocks! :rolleyes:
     
  7. macrumors 603

    VirtualRain

    Joined:
    Aug 1, 2008
    Location:
    Vancouver, BC
    #7
    If you ran a 10 minute encoding job, you wouldn't notice the 1 minute difference? :rolleyes:

    If you stripped out every advancement that only provided a 10% performance advantage over the years, where would you be? You can argue that it's not good value, but that's highly dependent on your workload and your budget.
     
  8. thread starter macrumors newbie

    Joined:
    Feb 27, 2009
    #8
    No, more like 5%-7%. Something you cannot really "feel". The reason I say the first benchmarks are not real world is because they are misleading. The actual difference is not 10% but more like 5% in real world use. And in terms of "feeling" this actual 5% difference it is even worse because you can't tell (like you can with some other upgrades).
     
  9. macrumors 604

    SnowLeopard2008

    Joined:
    Jul 4, 2008
    Location:
    Silicon Valley
    #9
    To summarize this thread is a few sentences, the difference is only felt if you are doing scientific research/calculations, simulations, or hard core CPU intensive tasks. For the average Pro user, higher clock speed makes a bigger difference. But for the ultimate Pro user, the number of cores makes a bigger difference.
     
  10. thread starter macrumors newbie

    Joined:
    Feb 27, 2009
    #10
    Here is a link to another test by MacWorld:

    http://www.macworld.com/article/139919/2009/04/cto_macpro.html

    It is not a fair test because the 2.93 Quad is using the faster 1TB WD Black hard drive as opposed to the slower 640GB WD Blue in the 2.66. The 2.93 is also outfitted with the ATI Radeon HD 4870 graphics card as opposed to the much slower Nvidia GeForce GT 120 graphics card.
     
  11. macrumors 601

    eXan

    Joined:
    Jan 10, 2005
    Location:
    Russia
    #11
    The Mac Pro CPU "upgrades" have always been worthless. Nothing new here.
     
  12. macrumors G4

    Joined:
    May 6, 2008
    #12
    Solving bottlenecks such as HDD throughput can make a significant difference. Both real world, and "felt" by the user. :D

    For the increase in cost for faster processors, it may be possible to implement a decent RAID setup. YMMV, depending on usage of course. :p
     
  13. thread starter macrumors newbie

    Joined:
    Feb 27, 2009
    #13
    This is getting funny so I'll summarize the whole thing...

    Cinebench and Geekbench say the 2.93 Quad is roughly 10% faster than the 2.66 Quad. Is this true? Yes. But what is the real world difference when using applications? 5%-7% Will I be able to notice this 5%-7% difference? I couldn't. What do you suggest I spend the $500 on instead? 3x 1TB WD Blacks + 6GB Ram, or $500 off a ACD, or SSD + 4870, etc. These are things you will really notice. What about a VelociRaptor? I wouldn't.
     
  14. macrumors regular

    Joined:
    Mar 18, 2003
    Location:
    Spring Hill, TN
    #14
    This is the reason I went with the 2.66 as well. I just couldn't justify the upgrade to the 2.93. I also got more ram and storage.
     
  15. macrumors 603

    VirtualRain

    Joined:
    Aug 1, 2008
    Location:
    Vancouver, BC
    #15
    EDIT: Upon some reflection, I retract these remarks. Everyone will and does place different value on this kind of thing... it was worth the $500 for me, but not for you. Fair enough.
     
  16. macrumors regular

    Joined:
    Jul 13, 2007
    #16
    I've run the digglloyd test in around 18 seconds on my machine.

    Here are my Cinebench scores.

    CINEBENCH R10
    ****************************************************

    Tester :

    Processor : Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz
    MHz : 3800
    Number of CPUs : 8
    Operating System : WINDOWS 64 BIT 6.0.6001

    Graphics Card : GeForce GTX 275/PCI/SSE2
    Resolution : <fill this out>
    Color Depth : <fill this out>

    ****************************************************

    Rendering (Single CPU): 5352 CB-CPU
    Rendering (Multiple CPU): 21949 CB-CPU

    Multiprocessor Speedup: 4.02

    Shading (OpenGL Standard) : 6800 CB-GFX


    ****************************************************
     
  17. macrumors 601

    Tesselator

    Joined:
    Jan 9, 2008
    Location:
    Japan
    #17
    How can you have 8 CPUs and only a 4X speed-up? (rhetorical) Those cores must REALLY SUCK nuggets.


    A+ for most accurate statement of the entire thread! ;)


    .
     
  18. macrumors regular

    Joined:
    Mar 28, 2009
    Location:
    Bavaria, Germany
    #18
    * 2.66 quad plus 20% more cash = 2.93 quad = kind of some more speed, you may not notice
    * 2.66 quad plus 14% more cash = 2.66 quad / Intel X25-M 80GB SSD = quite a lot of additional fun every day

    Agree?!
     
  19. macrumors G4

    Joined:
    May 6, 2008
    #19
    HDD performance is the single biggest problem for me, as the software I use is disk intensive. Graphics are nice, and CPU clocks are important as well, but if I can't get the data fed to the processor, it's speed doesn't make that much of a difference. ;)
     
  20. macrumors 6502

    Joined:
    May 24, 2007
    #20
    How much of a premium, if any, will someone pay for a 2.93 vs a 2.66 in two years? On eBay, Given two equal systems, save for processor speed... I think most people will choose the 2.93 over the 2.66...

    something to consider when unloading the machine in the future.
     
  21. macrumors 603

    VirtualRain

    Joined:
    Aug 1, 2008
    Location:
    Vancouver, BC
    #21
    * 2.66 quad plus 50% more cash = 2.93 quad / dual X25-M in RAID 0 = Priceless! :p :D
     
  22. macrumors 603

    VirtualRain

    Joined:
    Aug 1, 2008
    Location:
    Vancouver, BC
    #22
    Yes, I agree, one should always budget a portion of their computer spend on trying to optimize storage... great system performance is about balance.

    I think we all agree that this is sound investment strategy when it comes to PC's.
     
  23. macrumors 6502a

    Joined:
    Apr 22, 2009
    #23
    So if I buy a 2.93 Quad it comes with the better HD either WD Black (32MB cache) 640 or 1TB versus a WD Blue with 16MB cache in the 2.66 Quad?

     
  24. macrumors regular

    Joined:
    Mar 28, 2009
    Location:
    Bavaria, Germany
    #24
    How did I know YOU would add this option :D
    The better one is always the good one´s enemy...
    I don´t think so, but one with a 2.66 quad could tell us, please?! A Caviar Green system disk would be a no go for me.
    So: who went for a 2.66 quad and can tell us, what disk has been in there?! But as I know: apple changes, also.
     
  25. macrumors 603

    VirtualRain

    Joined:
    Aug 1, 2008
    Location:
    Vancouver, BC
    #25
    I believe if you opt for the 1TB HD (regardless of CPU choice) you will get a WD Black. If you opt for the standard 640GB disk, you may get either a WD Blue or Black (it's a lottery). Last I checked, WD does not make a 1TB Blue drive which is why the 1TB optional upgrade is a safe bet.

    :p
     

Share This Page