The Notebook Review forums were hosted by TechTarget, who shut down them down on January 31, 2022. This static read-only archive was pulled by NBR forum users between January 20 and January 31, 2022, in an effort to make sure that the valuable technical information that had been posted on the forums is preserved. For current discussions, many NBR forum users moved over to NotebookTalk.net after the shutdown.
Problems? See this thread at archive.org.
 Next page →

    Nvidia Quadro K2000M

    Discussion in 'Gaming (Software and Graphics Cards)' started by pterodactilo, Jun 5, 2012.

  1. pterodactilo

    pterodactilo Notebook Consultant

    Reputations:
    11
    Messages:
    225
    Likes Received:
    7
    Trophy Points:
    31
    Where can I find the specs of this card? laptops like Lenovo Thinkpad W530 that use it have been already released but I can't find any technical information on Nvidia webpage.
     
    pterodactilo, Jun 5, 2012
    #1
  2. Star Forge

    Star Forge Quaggan's Creed Redux!

    Reputations:
    1,676
    Messages:
    2,700
    Likes Received:
    10
    Trophy Points:
    56
    It is going to based around a crippled GTX 580M according to leaks from Lenovo China.
     
    Star Forge, Jun 5, 2012
    #2
  3. Meaker@Sager

    Meaker@Sager Company Representative

    Reputations:
    9,436
    Messages:
    58,194
    Likes Received:
    17,902
    Trophy Points:
    931
    2GB DDR3 means its likely based on the 650M chip.
     
    Meaker@Sager, Jun 5, 2012
    #3
  4. Star Forge

    Star Forge Quaggan's Creed Redux!

    Reputations:
    1,676
    Messages:
    2,700
    Likes Received:
    10
    Trophy Points:
    56
    Rumored to have 384 CUDA Cores. So either it is in Kepler figures (650M) or it is a crippled 580M.
     
    Star Forge, Jun 5, 2012
    #4
  5. Meaker@Sager

    Meaker@Sager Company Representative

    Reputations:
    9,436
    Messages:
    58,194
    Likes Received:
    17,902
    Trophy Points:
    931
    Very much doubt 580m.... with ddr3? nahhhh, that would be crazy power inefficient.
     
    Meaker@Sager, Jun 5, 2012
    #5
  6. pterodactilo

    pterodactilo Notebook Consultant

    Reputations:
    11
    Messages:
    225
    Likes Received:
    7
    Trophy Points:
    31
    I don't understand why Nvidia keeps using DDR3 in mobile mid-range professional cards. In theory the 28nm node with lower power consumption should give enough headroom for faster ram.
     
    pterodactilo, Jun 5, 2012
    #6
  7. sgogeta4

    sgogeta4 Notebook Nobel Laureate

    Reputations:
    2,389
    Messages:
    10,552
    Likes Received:
    7
    Trophy Points:
    456
    Because it's cheaper. That and the company using the gpu chooses it. They can easily use a gddr5 model if they want to.

    Sent from my Galaxy Nexus using Tapatalk
     
    sgogeta4, Jun 5, 2012
    #7
  8. Karamazovmm

    Karamazovmm Overthinking? Always!

    Reputations:
    2,365
    Messages:
    9,422
    Likes Received:
    200
    Trophy Points:
    231
    speed isnt the main thing, precision is. The price for GDDR5 ECC buffered is more expensive than the normal GDDR5
     
    Karamazovmm, Jun 5, 2012
    #8
  9. fzhfzh

    fzhfzh Notebook Deity

    Reputations:
    289
    Messages:
    1,588
    Likes Received:
    0
    Trophy Points:
    55
    I don't think they use ECC buffered vram for mobile quadros except for the top end mobile quadro.
     
    fzhfzh, Jun 5, 2012
    #9
  10. Karamazovmm

    Karamazovmm Overthinking? Always!

    Reputations:
    2,365
    Messages:
    9,422
    Likes Received:
    200
    Trophy Points:
    231
    they do use for all the quadros. the NVS are just with a different driver
     
    Karamazovmm, Jun 6, 2012
    #10
  11. thinkhd

    thinkhd Newbie

    Reputations:
    0
    Messages:
    1
    Likes Received:
    0
    Trophy Points:
    5
    thinkhd, Jun 6, 2012
    #11
  12. rashaanp

    rashaanp Newbie

    Reputations:
    0
    Messages:
    7
    Likes Received:
    0
    Trophy Points:
    5
    rashaanp, Jun 7, 2012
    #12
  13. Star Forge

    Star Forge Quaggan's Creed Redux!

    Reputations:
    1,676
    Messages:
    2,700
    Likes Received:
    10
    Trophy Points:
    56
    If it is the Kepler version, then it is weaker than its predecessor, the 2000M.
     
    Star Forge, Jun 7, 2012
    #13
  14. Meaker@Sager

    Meaker@Sager Company Representative

    Reputations:
    9,436
    Messages:
    58,194
    Likes Received:
    17,902
    Trophy Points:
    931
    It must be kepler with 29Gbps bandwidth. Also I think the k at the front is a givaway.
     
    Meaker@Sager, Jun 7, 2012
    #14
  15. grimace912

    grimace912 Notebook Enthusiast

    Reputations:
    0
    Messages:
    25
    Likes Received:
    0
    Trophy Points:
    5
    Doesn't the PDF mention that they're all Fermi on the second page?

    If it's Fermi with 384 cores... would that be better for gaming than a 650m?
     
    grimace912, Jun 7, 2012
    #15
  16. Star Forge

    Star Forge Quaggan's Creed Redux!

    Reputations:
    1,676
    Messages:
    2,700
    Likes Received:
    10
    Trophy Points:
    56
    Yeah you are right. Unless it is a typo, it is really a severely crippled 580M lol...
     
    Star Forge, Jun 7, 2012
    #16
  17. grimace912

    grimace912 Notebook Enthusiast

    Reputations:
    0
    Messages:
    25
    Likes Received:
    0
    Trophy Points:
    5
    Sorry I'm kinda really out of whack with mobile gpus... but would this be better than a kepler 650m? If so, by how much?
     
    grimace912, Jun 7, 2012
    #17
  18. Star Forge

    Star Forge Quaggan's Creed Redux!

    Reputations:
    1,676
    Messages:
    2,700
    Likes Received:
    10
    Trophy Points:
    56
    Well Kepler CUDA Cores is compared as 2 Kepler Cores to 1 Fermi Core roughly. So 384 Kepler Cores is like 192 Fermi Cores roughly, which would be around a GTX 560M-level card that sported GDDR3 (GTX 560 had 192 CUDA Cores). If the 384 CUDA Cores was Fermi-based however, then it would be around a GTX 580M (which had 384 CUDA Cores). However, since the GTX 580M was released with a 192-bit GDDR 5 interface and the K2000M in the old 128-bit GDDR3, then we are basically seeing a 580M crippled by its VRAM.
     
    Star Forge, Jun 7, 2012
    #18
  19. grimace912

    grimace912 Notebook Enthusiast

    Reputations:
    0
    Messages:
    25
    Likes Received:
    0
    Trophy Points:
    5
    Ah I see what you mean by it being a gimped 580M. It looks like the 650M uses either GDDR3 or GDDR5.

    So clearly the K2000M would crush a GDDR3 650M, right? How exactly would it stack up against a GDDR5 650M?
     
    grimace912, Jun 7, 2012
    #19
  20. Meaker@Sager

    Meaker@Sager Company Representative

    Reputations:
    9,436
    Messages:
    58,194
    Likes Received:
    17,902
    Trophy Points:
    931
    Noooooo, a 580M core with 128bit DDR3? I don't believe that....
     
    Meaker@Sager, Jun 7, 2012
    #20
  21. Star Forge

    Star Forge Quaggan's Creed Redux!

    Reputations:
    1,676
    Messages:
    2,700
    Likes Received:
    10
    Trophy Points:
    56
    Well nVidia has done weird things before so it might be true...
     
    Star Forge, Jun 7, 2012
    #21
  22. Syllogistic

    Syllogistic Notebook Enthusiast

    Reputations:
    19
    Messages:
    22
    Likes Received:
    0
    Trophy Points:
    5
    Would be really useful to have some reviews of this card before committing to a W530, but it doesn't look like it's even been officially announced by Nvidia yet? Very strange that there's so little information available. I would hope it performs better than the 650M.
     
    Syllogistic, Jun 7, 2012
    #22
  23. Star Forge

    Star Forge Quaggan's Creed Redux!

    Reputations:
    1,676
    Messages:
    2,700
    Likes Received:
    10
    Trophy Points:
    56
    You are crippled by the RAM Regardless. 128-bit GDDR3 was from like 2008 and it is starting to age fast. 128-bit GDDR5 is like the absolute minimum for good performing cards or at least 192-bit GDDR3. Even a 650M will see high-yields of VRAM bottlenecking with a total bandwidth of like 26 GB/s to go with.
     
    Star Forge, Jun 7, 2012
    #23
  24. Syllogistic

    Syllogistic Notebook Enthusiast

    Reputations:
    19
    Messages:
    22
    Likes Received:
    0
    Trophy Points:
    5
    Sadly there isn't really another option, I don't think. Nvidia seems to have dropped the ball with mobile (and arguably desktop) Kepler so far.
     
    Syllogistic, Jun 7, 2012
    #24
  25. Star Forge

    Star Forge Quaggan's Creed Redux!

    Reputations:
    1,676
    Messages:
    2,700
    Likes Received:
    10
    Trophy Points:
    56
    Desktop Kepler is very good but they can't keep up with demand though...
     
    Star Forge, Jun 7, 2012
    #25
  26. pterodactilo

    pterodactilo Notebook Consultant

    Reputations:
    11
    Messages:
    225
    Likes Received:
    7
    Trophy Points:
    31

    It makes sense because in professional applications, which is the market this card is targeted to, it's more important more cores than more bandwidth . Add to that the fact that Kepler is a worse architecture than Fermi for parallel computing. I'm nonetheless pretty much interested in the gaming performance only. :D
     
    pterodactilo, Jun 7, 2012
    #26
  27. Star Forge

    Star Forge Quaggan's Creed Redux!

    Reputations:
    1,676
    Messages:
    2,700
    Likes Received:
    10
    Trophy Points:
    56
    How is it worse for parallel architecture? Isn't it suppose to be better than Fermi in raw computational power as well?
     
    Star Forge, Jun 7, 2012
    #27
  28. Meaker@Sager

    Meaker@Sager Company Representative

    Reputations:
    9,436
    Messages:
    58,194
    Likes Received:
    17,902
    Trophy Points:
    931
    Not for CUDA due to lack of FP64 performance.
     
    Meaker@Sager, Jun 7, 2012
    #28
  29. KCETech1

    KCETech1 Notebook Prophet

    Reputations:
    2,527
    Messages:
    4,112
    Likes Received:
    449
    Trophy Points:
    151
    its slow as crap in OpenCL apps ( Adobe CS6 and MAYA ) one of my pro video editing apps uses fp64 and that moves it to even WORSE! Kepler arch seems to be nutered and designed purely for gaming and fp32 to compete in the gaming segment and in the process gave up GPGPU

    I was hoping quadros didnt have the nutered performance of the gforce cards but sofar reviews arent good at all and theyre staying with fermi to try and be semi competitive
     
    KCETech1, Jun 7, 2012
    #29
  30. Star Forge

    Star Forge Quaggan's Creed Redux!

    Reputations:
    1,676
    Messages:
    2,700
    Likes Received:
    10
    Trophy Points:
    56
    Hmmmm... that is a revelation indeed. I thought they would totally push Kepler on both fronts since they released the Tesla GK110 out just for the occasion but it looks like that either due to low supply or phasing out CUDA that they are going backwards for Kepler in professional applications.
     
    Star Forge, Jun 7, 2012
    #30
  31. Meaker@Sager

    Meaker@Sager Company Representative

    Reputations:
    9,436
    Messages:
    58,194
    Likes Received:
    17,902
    Trophy Points:
    931
    It's because they intended for big kepler (GK100) to be their flagship but it got canned because it was not producible.

    The refresh GK110 may only see use in tesler or may get released with the 7 series.
     
    Meaker@Sager, Jun 7, 2012
    #31
  32. Syllogistic

    Syllogistic Notebook Enthusiast

    Reputations:
    19
    Messages:
    22
    Likes Received:
    0
    Trophy Points:
    5
    Seems like NVIDIA have now quietly released some info on these cards -- the specs are in this comparison chart: http://www.nvidia.com/content/PDF/product-comparison/Quadro_Mobile_Product_Comparison.pdf

    So that confirms this is a card with 128-bit DDR3. Very disappointing. Seems like I would be better off buying a laptop with integrated graphics (eg X230, T530) and using an external GPU solution rather than shelling out for the W530 with K2000. Thoughts?
     
    Syllogistic, Jun 7, 2012
    #32
  33. pterodactilo

    pterodactilo Notebook Consultant

    Reputations:
    11
    Messages:
    225
    Likes Received:
    7
    Trophy Points:
    31
    What about other options from AMD? For example in HP elitebook 8570 and 8770 one can choose AMD Firepro Mobility 4000 as alternative to nvidia cards. It has GDDR5 as memory interface . Even if it is based on 7750M I guess that performace will be good in games.
     
    pterodactilo, Jun 7, 2012
    #33
  34. Star Forge

    Star Forge Quaggan's Creed Redux!

    Reputations:
    1,676
    Messages:
    2,700
    Likes Received:
    10
    Trophy Points:
    56
    You will be working with half-a**ed drivers but yeah the hardware will be faster. Even if the K2000M is Fermi-based 580M, the 128-bit GDDR3 will cripple it hard regardless. You are way better off with the AMD solutions.
     
    Star Forge, Jun 7, 2012
    #34
  35. pterodactilo

    pterodactilo Notebook Consultant

    Reputations:
    11
    Messages:
    225
    Likes Received:
    7
    Trophy Points:
    31
    Do you refer to standard drivers or drivers for professional applications and Linux support? Personally I'm only concerned about gaming performance on windows platform, so I guess that I can trade worse drivers for faster hardware. It depends on personal needs though. As for the K2000M, is it based on 28 nm fabrication process at least? I mean, all the Fermi based cards of the 6XX series like 670M are 40nm . Even if DDR3 consumes less than GDDR5, 382 Fermi cores are a lot of cores for a 55W card.
     
    pterodactilo, Jun 7, 2012
    #35
  36. Star Forge

    Star Forge Quaggan's Creed Redux!

    Reputations:
    1,676
    Messages:
    2,700
    Likes Received:
    10
    Trophy Points:
    56
    the Cata;yst drivers aren't really that good but hardware-wise you are better off I truly think. Regardless if it is Fermi or Kepler, the nVidia solution is just not strong.
     
    Star Forge, Jun 7, 2012
    #36
  37. Syllogistic

    Syllogistic Notebook Enthusiast

    Reputations:
    19
    Messages:
    22
    Likes Received:
    0
    Trophy Points:
    5
    The problem for me with the Dell and HP offerings, as I understand them, is that battery life is poor when you select the high-quality displays because the integrated graphics don't support 10-bit displays. The Lenovos have poorer but still very good quality displays that do support switchable graphics so you get a much better battery life.

    Frankly I'm not THAT concerned about gaming since I have a good desktop PC. But I'm expecting to travel a fair bit in the next couple of years so would like something I can at least decently game on while I'm away, as well as something with good battery life. Seem like pretty difficult criteria to meet at the moment!
     
    Syllogistic, Jun 7, 2012
    #37
  38. grimace912

    grimace912 Notebook Enthusiast

    Reputations:
    0
    Messages:
    25
    Likes Received:
    0
    Trophy Points:
    5
    grimace912, Jun 7, 2012
    #38
  39. Syllogistic

    Syllogistic Notebook Enthusiast

    Reputations:
    19
    Messages:
    22
    Likes Received:
    0
    Trophy Points:
    5
    I had assumed from the change in mobile platform generation, the number of CUDA cores and the "K" prefix, but in retrospect, none of those directly confirm it's Kepler. I'll amend my earlier post. I suppose we'll have to wait for a press release or for someone to actually get their hands on one to confirm.
     
    Syllogistic, Jun 7, 2012
    #39
  40. KCETech1

    KCETech1 Notebook Prophet

    Reputations:
    2,527
    Messages:
    4,112
    Likes Received:
    449
    Trophy Points:
    151
    KCETech1, Jun 7, 2012
    #40
  41. Syllogistic

    Syllogistic Notebook Enthusiast

    Reputations:
    19
    Messages:
    22
    Likes Received:
    0
    Trophy Points:
    5
    That doc actually says Fermi on p 2, doesn't it?
     
    Syllogistic, Jun 7, 2012
    #41
  42. pterodactilo

    pterodactilo Notebook Consultant

    Reputations:
    11
    Messages:
    225
    Likes Received:
    7
    Trophy Points:
    31
    I think it's based on Kepler. 384 fermi cores and 55W makes no sense.
     
    pterodactilo, Jun 7, 2012
    #42
  43. KCETech1

    KCETech1 Notebook Prophet

    Reputations:
    2,527
    Messages:
    4,112
    Likes Received:
    449
    Trophy Points:
    151
    yes it says Fermi in the document.
    Kepler makes no sense as the 104 core is crap for GPGPU and im thinking we need to wait fir the next chip

    dont think so the nvidia document linked above says its fermi on page 2 and not using kepler makes sense as fermi is faster in OpenCL, and FP64. kepler got neutered to compete in gaming and sacrificed GPGPU functions
     
    KCETech1, Jun 7, 2012
    #43
  44. grimace912

    grimace912 Notebook Enthusiast

    Reputations:
    0
    Messages:
    25
    Likes Received:
    0
    Trophy Points:
    5
    KCETech1 - Yeah I pointed out earlier in the thread that it said Fermi on that doc too. I was wondering how Syllogistic deduced that the other document confirmed that the gpu was Kepler.

    Regardless, I'm pretty sure Nvidia said Quadro Kepler wasn't coming until next spring or something...
     
    grimace912, Jun 7, 2012
    #44
  45. Syllogistic

    Syllogistic Notebook Enthusiast

    Reputations:
    19
    Messages:
    22
    Likes Received:
    0
    Trophy Points:
    5
    As others have said, it just seemed to make more sense for it to be Kepler given the new generation, "K", TDP vs CUDA cores etc, but it seems like it would be a good thing if I was proven wrong so I hope that's the case!
     
    Syllogistic, Jun 7, 2012
    #45
  46. grimace912

    grimace912 Notebook Enthusiast

    Reputations:
    0
    Messages:
    25
    Likes Received:
    0
    Trophy Points:
    5
    So here ( Kepler GPU hard launch now expected in April

    "The company originally scheduled GeForce Kepler for the launch in the first quarter (i.e. fiscal fourth quarter 2012), Quadro Kepler was scheduled for the second quarter (i.e. fiscal first quarter 2013) and Tesla Kepler was scheduled for the third quarter (second fiscal quarter 2013)."
     
    grimace912, Jun 7, 2012
    #46
  47. Carcozep

    Carcozep Notebook Enthusiast

    Reputations:
    66
    Messages:
    43
    Likes Received:
    76
    Trophy Points:
    26
    I want to come and help this discussion since I have one last hope for Nvidia on doing something proper. They have disappointed me with the 660M with the minimal increase over 560M while keeping the same TDP.

    From the 2nd document posted here, I believe we are dealing with a Kepler based Quadro, coming from the big, GK110 Tesla Die. A 384 shader Fermi, to fit in 55W, would mean really really low clock speeds. Somewhere in the realm of 300-400 Mhz. Performance would be downright bad.

    So, after fiddling with the big Tesla Die, I believe this is the most probable outcome of the K2000M:
    [​IMG]

    Although it says Pci-Ex 2 in the Nvidia Paper, I didn't manage to find a proper header like this one with the Pci-Ex 3. I did my best with the limited functionality of Paint. The small yellow blocks would be the Double Precision Units. It appears Nvidia went on the route of making one architecture for gaming, and one for computing. If what I assume is true, they would keep those units in the mobile Quadros and not give us a gaming GK107 with some optimized drivers.

    On a more pessimistic closing note, Why did they not cram 768 shaders into 75W TDP? They managed to squeeze 1344 in the realm of 100+W TDP, the 680M. :(
     
    Carcozep, Jun 8, 2012
    #47
  48. pterodactilo

    pterodactilo Notebook Consultant

    Reputations:
    11
    Messages:
    225
    Likes Received:
    7
    Trophy Points:
    31
    There is no FP64 precission support in K2000M. Look at this table:
    http://www.nvidia.com/content/PDF/product-comparison/Quadro_Mobile_Product_Comparison.pdf
     
    pterodactilo, Jun 8, 2012
    #48
  49. KCETech1

    KCETech1 Notebook Prophet

    Reputations:
    2,527
    Messages:
    4,112
    Likes Received:
    449
    Trophy Points:
    151
    KCETech1, Jun 8, 2012
    #49
  50. Carcozep

    Carcozep Notebook Enthusiast

    Reputations:
    66
    Messages:
    43
    Likes Received:
    76
    Trophy Points:
    26
    Well, if that holds true and there are no double precision units, we shall receive the gaming GK107 and GK104 as Quadro's. Now that future is really grim for mobile computing :(
     
    Carcozep, Jun 8, 2012
    #50
 Next page →