|
Post by The Guy on Mar 20, 2010 14:50:11 GMT
As everyone knows ATI dominated 2009 and the exciting battle continues to 2010. Many electronics such as the mobile products this year is going be most epic. AMD has released its evergreen on time but Nvidia has officially pushed Fermi to march. Most likely fermi>evergreen , but IMO the HD6xxx will be EPIC.
I am holding my breath till the HD6000 and I want to know which company do you think will dominate and give conclusive reason and argument to it.
|
|
|
Post by cew27 on Mar 20, 2010 14:53:52 GMT
right now ATI hold both the performance AND price/performance crown. normally they duke it out, but fermi is delayed and given ATI a chance to pull ahead.
the way i see it, is that ATI have done something amazing with the 5 series cards. the performance jump between this and the last gen is absolutely huge. much more than most generation leaps. Nvidia need to really come up with something good to come out on top. and even when they do, it wont be cheap. i really dont think nvidia have it in them this gen. and i think they will be playing catchup for the rest of the year.
even if the extra time on fermi allows it to rival the 5870 (i doubt they could surpass it, given the massive performance leap) it will be so much more expensive by the time its released, that only die hard nvidsia fans (fanbois?) will buy them.
ive got nothing against nvidiam my alegience goes with whoever currently offers thebest cards at the best price. right now, that ATI without a doubt. and i have a feeling it will stay that way.
but if nvidia do come out with sometihng more powerful at a reasonable cost, ill surely be getting one instead of a 5870. but right now, im expecting to be purchasing a 5870.
|
|
|
Post by rakar on Mar 20, 2010 14:59:13 GMT
If anything, I only see ATi's lead growing through 2010.
*Assuming* Fermi does come out in March/April... there are a couple of fundamental issues Nvidia have to overcome:
1. It is not competitive per mm^2 of chip compared with Evergreen. 2. Even the top of the line Fermi is not going to be quicker than Hemlock.
Thus, ATi can set the prices from high end, and can scale them appropriately to ensure they (ATi) make a profit, and Nvidia cannot.
Two other issues are:
1. With Nvidia's design team spending so long on Fermi, how can they expect to catch up in 1 generation? Can Nvidia's finances afford them to take several generations to catch up?
2. Nvidia have to compromise the graphics performance for compute performance. ATi can offload this to their (AMD) x86 CPUs... indeed, the bulldozer archiecture is the first step toward designing CPUs and GPUs as one integrated system which load-balances depending on FP/Int requirements. This is a fundamental problem with their (Nvidia's) strategic direction that they (Nvidia) need to sort - otherwise they will never be competitive with ATi on a per mm^2 basis.
|
|