NV40 Under the Microscope


The NV40 chip itself is massive. Weighing in at a hefty 222 Million transistors, NVIDIA's newest GPU has more than three times the number of transistors as Intel's Northwood P4, and about 33% more transistors than the Pentium 4 EE. This die is droped onto a 40mm x 40mm flipchip BGA package.



NVIDIA doesn't publish their die size information, but we have been able to interpolate a little bit from the data we have available on their process information, and a very useful wafer shot.



As we can see, somewhere around 16 chips fit horizontally on the wafer, while they can squeeze in about 18 chips vertically. We know that NVIDIA uses a 130nm IBM process on 300mm wafers. We also know that the P4 EE is in the neighborhood of 250mm^2 in size. Doing the math indicates that the NV40 GPU is somewhere between 270mm^2 and 305mm^2. It is difficult to get a closer estimate because we don't know how much space is between each chip on the wafer (which also makes it hard to estimate waste per wafer).

Since we don't have information on yields either, it's hard to say how well NVIDIA will be making out on this GPU. Increasing the transistor count and die size will lower yields, and the retail value of cards based on NV40 will have the same price at release as when NV38 was released.

Of course, even if they don't end up making as much money as they want off of this card, throwing down the gauntlet and pushing everything as hard as they can will be worth it. After the GeForce FX series of cards failed to measure up to the hype (and the competition), NVIDIA has needed something to reestablish their position as performance leader in the industry. This industry can be brutal, and falling short twice is well nigh a death sentence.

But, all those transistors on such a big die must draw a lot of power right? Just how much juice do we need to feed this beast ...
What’s new in DX 9.0c Power Requirements
Comments Locked

77 Comments

View All Comments

  • Da3dalus - Thursday, April 15, 2004 - link

    I'd like to see benchmarks of Painkiller in the upcoming NV40 vs R420 tests...
  • Brickster - Thursday, April 15, 2004 - link

    Am I the only one who thinks Nvidia's Nalu is the MOST bone-able cartoon out there?

    Oy, get the KY!
  • Warder45 - Thursday, April 15, 2004 - link

    Did any reviews try and overclock the card? Is it not possible with the test card?
  • DonB - Thursday, April 15, 2004 - link

    Would have been better if it had a coax cable TV input + TV tuner. For $500, I would expect a graphic card to include EVERYTHING imaginable.
  • Pete - Thursday, April 15, 2004 - link

    Shinei #37,

    "Speaking of DX9/PS2.0, what about a Max Payne 2 benchmark?"

    MP2 doesn't use DX9 effects. The game requires DX9 compatability, but only DX8 compliance for full effects.

    Xbit-Labs has a ton of benches of next-gen titles as well, and is worth checking out. NV40 certainly redeems itself in the HL2 leak. :)
  • Wwhat - Thursday, April 15, 2004 - link

    Anybody happen to know if it's possible to use a second (old) PSU to run it, you can pick up cheap 235 watt PSU's and would be helped with both extra connectors and power.
    I'm not sure it won't cause 'sync' problems though as a small difference between the rails of 2 PSU's would cause one to drain the other if the card's connectors aren't decoupled enough from the AGP port.



  • Pumpkinierre - Thursday, April 15, 2004 - link

    Agrre with you Trog #59 on the venting. Also with DX9.0c having fp32 as spec., does this mean that FX series cards redeem themselves? (As the earlier DX9 spec was fp24 which was'nt present on the FX gpus causing a juggling act between fp16 and fp32 to match performance and IQ). Still, full fp32 on the FX cards might be too slow.
  • mrprotagonist - Thursday, April 15, 2004 - link

    What's with all the cheesy comments before the benchmarks? Anyone?
  • Cygni - Thursday, April 15, 2004 - link

    "what mobo and mobo drivers were used? i hear that the nforce2 provides an unfair performance advantage for nvidia"

    The test was on an Athlon 64 3400+ system, so i doubt it was using an Nforce2. But ya, i agree, the system specs were short. More details are required.
  • Brickster - Wednesday, April 14, 2004 - link

    Derek, what was that Monitor you used?

    Thanks!

Log in

Don't have an account? Sign up now