Intel Graphics HD3000 vs NVIDIA GeForce 320M/330M GT - Real Benchmarks and Comparison

Discussion in 'MacBook Pro' started by AppleMacFinder, Feb 25, 2011.

  1. macrumors 6502a

    AppleMacFinder

    Joined:
    Dec 7, 2009
    #1
    [​IMG]

    All the benchmarks were taken from www.notebookcheck.net.
    The results are quite reliable, I have used this source for years.

    According to benchmarks, HD3000 is BETTER than 320M !

    Also, remember that 320M has 256MB of shared memory,
    while HD3000 has 384MB - 1.5x more memory for game textures!

    But it is worse than 330M GT, if don't count the Cinebench.

    Hope this information will help you to make a right choice. :rolleyes:
     
  2. macrumors 6502

    Joined:
    Sep 30, 2008
    #2
    The Theoretical output is technically faster, but you have to one figure that game code has been designed with gpu's like the 320/330m in mine and that Intel obviously is gonna have to work on drivers.
     
  3. macrumors 6502a

    Joined:
    Jun 22, 2010
    #3
    by your logic the HD 3000 should be even faster than what those benchmarks show, since Nvidia's drivers are probably much better than intel's.
     
  4. macrumors 65816

    mgartner0622

    Joined:
    Jun 6, 2010
    Location:
    Colorado, USA
    #4
    Yeah, Intel's last couple of cards were pretty terrible when they first came out, although driver improvements slightly made them better.
    Hopefully we see some Apple Driver Updates for these new Intel Cards. You never know. :cool:
     
  5. macrumors 6502

    Joined:
    May 20, 2010
    #5
    A lot of these tests give the grade based on cpu and gpu, so it isn't a great direct comparison for just video performance. That and the 320m will naturally have more mature drivers.
     
  6. macrumors 65816

    dagamer34

    Joined:
    May 1, 2007
    Location:
    Houston, TX
    #6
    It's not like you can separate the Intel HD GPU from the same die the processors are on. If you have a Mac with a 320m, it has a Core 2 Duo. If you have a Mac with a Intel HD 3000, it has a Core i3/i5/i7.

    Comparing video performance only would be useful if it were a desktop GPU that could be removed. It's not.

    And more mature nVidia drivers should favor the 320M to be faster, not slower.
     
  7. macrumors regular

    FriarNurgle

    Joined:
    Jan 2, 2011
    #7
    So the HD3000 could potentially get better.

    I've been town between the old (refurbs) and the new 13" MBP. The price difference just isn't that great even between the old higher model and base new model. So the potential of increase in performance in the newer model has me leaning towards a new one... but then I see the benchmarks for the new 15" models and start questioning myself and looking for change in the couch cushions.
     
  8. macrumors 6502

    Joined:
    Jan 21, 2011
    Location:
    Adelaide, SA
    #8
    notebookcheck is a good source. I expect the 320M to be similar to the HD3000, and I expect the GT 330M to be similar to the 6490M.
     
  9. macrumors 6502

    Joined:
    May 20, 2010
    #9
    Yes and no. The cpu performance won't always directly dictate FPS in a game, but it will dictate a higher score in those benchmarks. A better comparison would honestly be some side by side videos of games running on both systems.
     
  10. thread starter macrumors 6502a

    AppleMacFinder

    Joined:
    Dec 7, 2009
    #10
    Last gen

    Yes, all the previous generations of Intel Graphics sucked badly.
    But the last one is acceptable, thanks to the architecture improvements.
     
  11. thread starter macrumors 6502a

    AppleMacFinder

    Joined:
    Dec 7, 2009
    #11
    Look at the right side of a table - there are FPS benchmarks.
     
  12. macrumors 65816

    henrikrox

    Joined:
    Feb 3, 2010
    #12
    Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; nb-no) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

    Epic fail. In the links you show you see the Intel 3000 being very slow in games like black ops. The intel 3000 hd can't play anything on medium settings.

    Stop saying this is real benchmarks. You are just posting from a site.

    And what you didn't think about Is that in the link you posted the intel igp is running along side a high end quad core mobile sandy bridge CPU. What do you think happens when you don't have that CPU and replace with a mid end i5 dual core which you find in the 13.

    Learn to read the facts
     
  13. macrumors 65816

    dagamer34

    Joined:
    May 1, 2007
    Location:
    Houston, TX
    #13
    I was looking at the game benchmarks which showed more FPS. That's the desirable end result of any CPU/GPU changes.
     
  14. macrumors 65816

    henrikrox

    Joined:
    Feb 3, 2010
    #14
    Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; nb-no) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

    You can't just look at numbers and expect something that is fail 101.

    No the intel 3000 hd is not as good as the 320m. Already been many 3d mark tests to prove that.

    Not he and 6470 is not as good as the gt330m.

    You can't just expect new stuff to be better
     
  15. AppleMacFinder, Feb 25, 2011
    Last edited: Feb 25, 2011

    thread starter macrumors 6502a

    AppleMacFinder

    Joined:
    Dec 7, 2009
    #15
    And where are the links? :rolleyes:
     
  16. macrumors 6502

    Joined:
    May 20, 2010
    #16
    For 2 games, one which is old. This is a pretty narrow picture, and doesn't list the specific settings used. There are tons of game companies who use different game engines and what not. If you only play those 2 games you are set I guess?
     
  17. macrumors 6502

    Joined:
    Jan 21, 2011
    Location:
    Adelaide, SA
    #17
    Nice job totally assuming I'm retarded. Fortunately I don't just expect new stuff to be better, nor did I say it would be.

    The question is harder than that. The HD 3000 is much further integrated than any previous IGP. It is power gated and monitored by SB's management subsystem. The fast L3 is shared with the GPU along the ring bus, as is the ability to turbo based on current TDP. Memory access is also faster due to the DMI bus, and you get more of it. Even if the HD 3000 has lower raw processing power, it has a lots of things working in its favour. Most of the problems we are seeing with it right now in games and at higher graphics settings are driver issues. These will obviously be cleaned up slowly as the drivers mature, and games begin to be tested against the new hardware. The potential for the HD 3000 to be as fast or faster than the 320M is definitely there on paper and in practice.

    As for the GT 330M vs the 6490M, there are a lot more knowns in that arena. The memory bandwidth is identical, and the 6490M is faster in raw flops. The TDP of both chips are quite similar. Whether or not that materializes into performing in a very similar way to the GT 330M will depend on drivers as well. How Apple's implementation of Nvidia's drivers compare to AMD's. Overall I expect the new chip to be a bit slower, but not tremendously so.


    By and large I feel like we got a side-grade in both sectors. I doubt any of the new base model stuff is going to be charging ahead, nor lagging behind too greatly. We have to remember that in the case of the HD 3000 this is Rev A hardware, it is new.
     
  18. macrumors 68000

    Littleodie914

    Joined:
    Jun 9, 2004
    Location:
    Rochester, NY
    #18
    Well the OP posted links to a site with actual benchmark results demonstrating that (in ALL 4 3DMark versions) the HD 3000 actually is faster. Until you can provide sources stating otherwise, it's your word against his actual proof.
     
  19. macrumors 65816

    henrikrox

    Joined:
    Feb 3, 2010
    #19
    Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; nb-no) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

    3dmark 06 1280×800

    2011 MBP with Intel HD 3000 = 4629
    2010 MBP with nVidia 320m = 4754
    Left 4 Dead – 1280×800 Med Settings (click here to see details of settings)

    2011 MBP with Intel HD 3000 (min/max/avg) = 38 / 90 / 63
    2010 MBP with nVidia 320m (min/max/avg) = 53 / 92 / 75
    Starcraft 2

    2011 MBP with Intel HD 3000 (min/max/avg) = HOLD until 2/25 1pm
    2010 MBP with nVidia 320m (min/max/avg) = HOLD until 2/25 1pm

    Please understand this. The link provided is with a quad core high end CPU along with th intel 3000 hd. There has already been 3d mark
    Posted by users here already

    The 2010 mbp with a core 2 duo 2,4ghz and a 320m scored 4700 and the 2011 i5 mbp with intel igp scored 4600

    Now you see how bad the gpu is? Because 3dmark also factor in CPU speed. And the i5 is clearly better then a core 2 duo.

    Still it lost. Ill find a link to shut everyone up. Saying the 3000 hd is good.
     
  20. macrumors 65816

    henrikrox

    Joined:
    Feb 3, 2010
    #20
    Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; nb-no) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

    http://www.techyalert.com/2011/02/25/macbook-pro-2010-vs-macbook-pro-2011/

    Here's the link. 12 fps less fps on average less compares to core 2 suo with a 320m. Even with a much faster CPU.

    Can we stop staying the intel gpu is as good
     
  21. thread starter macrumors 6502a

    AppleMacFinder

    Joined:
    Dec 7, 2009
    #21
    Benchmarks

    My table is based on 314 benchmarks.
    It is far more accurate than a couple of results. ;)
     
  22. macrumors 65816

    henrikrox

    Joined:
    Feb 3, 2010
    #22
    Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; nb-no) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

    Like synthetic benchmarks means something.
     
  23. thread starter macrumors 6502a

    AppleMacFinder

    Joined:
    Dec 7, 2009
    #23
    There are benchmarks for two games:
    Doom and Crysis, measured in FPS - it's not synthetic.
    HD3000 performed faster.
     
  24. macrumors 65816

    henrikrox

    Joined:
    Feb 3, 2010
    #24
    Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; nb-no) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

    There are for black ops aswel. There are for left for dead aswel.

    Core 2 duo with a 320m scored 50% higher fps then a intel 300 hd


    3dmark 06 1280×800

    2011 MBP with Intel HD 3000 = 4629
    2010 MBP with nVidia 320m = 4754


    It's all in the link I posted ;) love hearing about all the people who bought the new 13 defending the intel 300"hd. But thy will be disappointed
     
  25. thread starter macrumors 6502a

    AppleMacFinder

    Joined:
    Dec 7, 2009
    #25
    Black Ops and Left for Dead benchmarks based on a couple of results.
    Doom 3 and Crysis benchmarks based on more than three hundred.

    P.S. I do not have a 13" 2011. Nor I am going to buy one. :rolleyes:
     

Share This Page