i would take the opposite stance b/c right now.. like it or not ati is losing out to nviida.
so if your losing.. you NEED to you get something out that can compete w/ SLI and pixel shader 3.0.
if you delay then you only allow nvidia to ship more cards, earn more market share, make more profits, and gain mindshare.
so what ati needs to do is get the r520 out the door asap but if its delayed , and i think it will be due to SLI and the fact that ati's driver team is probably having a nightmare doing it, then the company is in trouble.
w/ g70 already taped out and ready to rock , if nvidia launches it first then its a double wammy against ati.
right now nvidia is in the position to just wait and see since , like i said.. like it or not they are in the lead w/ pixel shader 3.0 and SLI.
so they can sit back and just wait ati out and see what they have... NOT the opposite.
"so what ati needs to do is get the r520 out the door asap but if its delayed , and i think it will be due to SLI and the fact that ati's driver team is probably having a nightmare doing it, then the company is in trouble."
ATI needs to maximize return for its shareholders. Period. Sometimes that means delaying a product ready to be released if more profits can be made by doing so. If ATI feels it can make more money by release R520 at a later date, then that's the correct decision. Unless, of course, they feel they can put a nail in the coffin of NVDA by releasing it sooner rather than later. I happen to think both companies will be around for quite a while and realistically speaking, releasing R520 now might only kill sales of our top-line (high margin) chips.
We shouldn't worry about who's got the bigger dick, but how best to use it.
I think an important question is when R520 can/is available for shipping in quantity?
If it wouldn't have been available until the autumn/Sept. anyway then is it really a big difference?
Would agree, and as suggested in the article. This is simply tactical / strategic. Why introduce a new generation chip if they know/feel that current sales are doing well?
Of course, some will project onto this that the R520 is encountering delays b/c of problems being faced.
As far as who needs to make a move? Probably depends on your perspective.
Saying NVDA has the lead because of SLI is nonsense as it is like comparing apples to oranges. Single card leader is ATI. Who has been taking market share overall (with the exception of Intel)...ATI. Who's sales have been stagnating vs. who's has been growing?
Not to mention SLI is far from being free of bugs and the driver issue (per munki) seems to be a big weakness and it is yet to see if ATYT can come out with something better.
Not to mention that ATI's version will be coming out soon no doubt, probably irrespective of next gen. offerings.
The SM 3.0 argument seems to be a valid point, but how robust is NVDA's tech. in that regard? and everybody knows SM 3.0 will be coming with ATYT's next generation offerings.
Methinks it is NVDA who needs to be making the move...NOT ATI.
BUT...methinks NVDA can ill afford to have their next products lag too far behind ATI in this next round. That's why even though they probably should release in advance of ATYT (assuming they could which doesn't even seem to be the case, that could be testimony to problems they face or their technology), they will probably want to come out after so that either...:
1. They will know for sure if they have a 'faster' and/or competitive enough offering that is ok to be released.
2. They will go into damage control mode and try and beef up their offering as much as possible in order to compete with ATI prior to be released.