I thought that this post by WaltC over at Beyond3D was qiite a good read:
<< The problem is two-fold. First, I think we can dispense
withcalling these individuals "journalists"--they are not
journalists--they are individuals who run web sites who
thinkthat repeating unsubstantiated rumor is
"journalism." They simply do not understand
journalisticterms like "corroboration" or "verification"
or"attribution." Calling them "journalists" is extremely
charitableas well as wholly inaccurate.
The second problem although widespread is not so
easilyunderstandable. It lies in the individuals who read
unsubstantiatedrumor and believe it immediately even
thoughit is wholly unsubstantiated.
For instance, in this specific example, it is purported that
3dfxspent "$100 million" on the development of the V5
6K.Now, of course, that's laughable and a dead
giveawaythat the comments are utterly bogus. But it
doespoint out the problem that many are completely
unableto see how ridiculous this statement is at all
levels,presumably due to a thorough lack of fiscal
understanding.
So we have the worst possible combination: ignorant
webmastersspreading ignorant rumors to ignorant
people... Of necessity, then, it is incumbent on 3dfx
toconsistentlydeliver a PR stream of information which
nullifiesthe effects of persistent and inflammatory
rumor.To some extent the rumors fill a void created by
thelack of 3dfx response to legitimate questions, such
as:"What happened to Rampage?"
According to comments made by Greg Ballard a year
agoprior to his departure, Rampage was scheduled for
afall 2K ship date at the very latest. His comments
alludedto some $10 million in R&D that had been
investedin Rampage up to that point in time. Flash
forwardto where we are now, a year later, and it is not
presumptuousto estimate that 3dfx has at least $15
millionin R&D invested into the Rampage designs as well
as,according to Ballard's comments last year, about
threeyears' time invested into the Rampage project.
(Ballardsaid last year that the Rampage project had
alreadybeen on the drawing boards for 22 months.)
The reality is that if 3dfx cannot execute Rampage then
thecompany does itself no service by withholding this
infofrom the public and investors. This is so because
withoutRampage 3dfx will have no new products to
offerfor most of next year, which puts the survival of
thecompany in question. The situation resulting is a
downwardspiral of revenue based on a lack of new,
competitiveproducts to sell into the markets, which is
furtherinflammed by an ever-decreasing sell-through
ofthe existing product lines due to competitive
pressuresfrom other companies who are shipping new
products(eg, nVidia and ATI.) This in turn results in
lessand less money available for R&D and means that
thelikelihood of 3dfx actually being able to compete
diminishes.
In other words, 3dfx gains nothing financially by killing
Rampageand saying nothing about it as opposed to
killingit and saying something about it. Either way, the
company'sprospects would be equally dim.
What 3dfx desperately needs is a new-product
announcement!Of course, such an announcement
needsto based on products 3dfx can actually execute
withina reasonable amount of time post their
announcement.Saying nothing about anything they're
preparingto do leads to endless negative speculation
whichcan only hurt them, even if the speculation is
entirelywithout merit, as seems clear in this case.
The fact is that it has been almost two years since the
V3,and in that time 3dfx has produced a single chip
design--one--the VSA-100. Look at the number of chips
producedby its competitors since the V3, and the
reasonfor 3dfx's perdicament becomes obvious.
My own opinion is that without competitive products to
shipinto the market it really doesn't matter what 3dfx
sayspublicly at this point--the company has no future
withoutcompetitive products regardless of what it may
sayin a various PR inititatives.
On the other hand, if Rampage is 80-90% complete at
thispoint, which one would certainly hope to be the
case,then 3dfx has utterly failed by not talking about it
andannouncing it since Comdex 2K.
One other related point that comes to mind about 3dfx's
perceiveddirections is that 3dfx unfortunately seems to
believethat it can prosper on a "brand-related" level
whilesaying next to nothing about specific upcoming
products.This is very similar to what we've seen from
Applecomputer over the last couple of years, and I
thinkthis approach is a huge mistake. This kind of
approachis more or less a "style-over-substance"
campaignwhich is destined to fail, I believe. The reason
whysuch PR directions are destined to fail in the
hardwaretechnology sectors, whether you talk about
Appleor about 3dfx, is because, ultimately, people buy
hardwarefor its substance, and any positive response
frommarkets about style is only a temporary
phenomenonwhich the market will tire of just like any
sortof fad you can think of.
Steve Jobs at Apple is being forced to start talking
aboutthe substance of the hardware in Apple
computersthese days (from processors to CD-RW's,
youname it) because the Apple customer base is
becomingbored with loud colors and artistically shaped
plastic.While Apple no doubt considers the Cube an
artistictriumph, the markets have all but rejected it
completely,and the flop of the Cube is in no small
measureresponsible for the upcoming quarter-billion
dollarloss Apple is preparing to announce. The "reality
distortionfield" at Apple is even now fizzling out.
Should 3dfx take a similar route and decide to rely on
its"name" instead of on the products it sells, on the
hardwareit sells, it will find that its "name" has a very
limitedvalue in the longer run of things. V5 5.5K/V3
saleswill not suffice for most of next year.
And so, to the extent that 3dfx refrains from both
announcingand shipping new, competitive products
intothe market, the sort of mindless gossip sprouting
throughthe Internet is in part directly due to
negligenceat 3dfx, IMHO.
Of course, if you have nothing new remotely ready to
shipthere's no point in saying anything about it. I think
allof us sincerely hope that this is not the case, and that
3dfx'ssilence is due to errors of judgement in its public
relationsefforts instead of due to a complete lack of
havinganything under development which is close to
shipment. >>
As for losing respect for nVidia, they're profits are up 104% from last year (read at Shugashack), and they're products are in every major computer-maker's systems. They have taken the OEMl crown from ATi (all in 2 years).
They are in the mobile market with GeForce Go , have the X-Box deal, and will be producing mainboard chipsets.
I've been nothing but happy with their chips (Elsa Erazor TNT, Gigabyte SDR GeForce 256, Asus V7700 GeForce GTS). But I've also been happy with my 3dfx products (Canopus Pure3D, Voodoo 3 3000 AGP).
If anything, nVidia should be applauded for their rags to riches story...