Pros/Cons to 3dfx [long]:
I just bought into 3dfx since I view its "warning of great revenue"
to be the calm of this tech storm. But longer term, I think it might
be a great stock well into 1998. I've summarized some of my thoughts below ...
Pros
3dfx has the best 3D game-oriented graphics chipset on the market,
period. Anyone that has seen 3dfx boards knows this is true. And
the consensus is pretty universal among the gaming community.
While Riva-128 has done a good job in trying to match the speed
(but not quality) of the older Voodoo 1 chipset graphics, by the time
the competition has met Voodoo 1, we are about to see the raising
of the bar by another factor of 2x or 3x through Voodoo 2.
Getting game developers to write for your hardware is no small
task but it looks like 3dfx has made great strides over the last
year. By doing so, they are making their chipsets more
"proprietary." Also, there is substantial name recognition by
gamers of the 3dfx logo and quality.
The recent press releases shows that Voodoo demand is fantastic
and the company was supply not demand constrained. This will
help fuel the fire of developer support because if they can meet
all demand, the 3dfx chipsets will become the de facto gaming
target.
Looking at the earnings pre-release and their past quarter numbers,
it looks like there may be some possible upside to the current
consensus estimates of $.08. I am reluctant to reinforce any
absurd whisper numbers, but 3dfx is calling for a greater than
80% revenue increase from Q3.
Item Q3 Q4 baseline est.
Revenues 10,018 18,032
Gross Profit 4,666 8,114 [assuming same margins]
R&D 3,201 4,274
SG&A 2,684 2,857
Net Profit (872) 983
At approximately 12,454,000 shares, that puts us out at about the
$.08 quarter eps. R&D and SG&A was computed using the same
% sequential increase as from Q2 to Q3. Two big positives
here: (1) As volume increases, margins should get better. This is
particularly true if yields are improving. Someone with better
semi process experience might clarify this. The model, though,
only assumes 46.6% gross margins equal with Q3. At higher
volumes, I'd expect margins to improve due to decreased % of
fixed costs ... although it may be offset with lower pricing
for better market share penetration. (2) The release had said
3dfx would "at least double the Company's earlier estimate
of 40% revenue growth over the third quarter." I like the
"at least" part since the $18 million revenue figure is
hopefully higher.
The Banshee will really let 3dfx compete in the cost arena as well
as the quality arena. Right now, 3dfx is making great margins
because they have quality locked-up and people are willing to pay
a premium. This will likely continue but with Voodoo 2 chipsets
taking that position soon. But with the Banshee, 3dfx will be
able to move into the more traditional graphics markets like
those held by low-cost high volume producers like S3 and
currently targeted by Intel (760), Nvidia (Riva-128), and 3D
labs (Permedia 2). One could see their strategy of high-margin
products at the high-end which eventually get moved down
for lower-margin high-volume sales at the low-end.
Here's some food for thought: 3dfx lost a lot of stock price when
Sega dropped them from their game console. With the huge push
to drive down the cost of PCs and make them household
commodities, where does that leave 3dfx? What happens if the
whole PC becomes nothing more than a CPU, a DVD/CD-ROM, a cheap
removable high-speed drive, telephone/cable modem, and a 3dfx
Banshee? The game companies which write for PCs will have, for
the first time, a relatively low-cost platform that might compete
as a really high-end super Sega station but with internet
capabilities, disk storage, and all other PC applications.
If Tommy wants a low-end PC, what do you think he'll get?
A $300 PC with lame graphics or a $350 PC which can be a great
gaming system as well? The way the pricing is going, I think
such a scenario is plausible for 98 and certainly for 99.
Another thought. Jon Peddie & Associates does a number of 3D
graphics chip surveys. In a previous survey I read, they
mentioned that it will be a lot easier for companies with high-end
3D to move down to 2D than for 2D companies to move to high-end 3D.
We are seeing this played out with S3 that literally owned this
market 2 years ago. What about partnering or getting merged with
that big dinosaur S3? S3 has the most to gain and has a large % of
a nice big fab. S3 deals with volume productions. 3dfx has great
tech. Too bad S3 doesn't have the cash to even consider buying
3dfx.
Cons
Competition is tough. Nvidia is a good hardware company but they
look to be a little less than one-generation behind 3dfx. Intel,
while a real powerhouse, has used the Martin-Marietta Real3D chips
as their starting point for the i760. The Real3D chips have been
continuously delayed and preliminary reports say its (1) hot (as
in heat), and (2) like a Riva-128 in speed. Rendition keeps
delaying their next gen. S3 I think is toast. 3D labs is a
little behind Nvidia for the consumer market I think. And the
bottom line is that the initial trials of Voodoo 2 have blown away
the development community. Benchmarks of Quake with Voodoo 2
far exceed the Riva-128 (2x-3x) and the Riva has poorer quality.
But competition is the key and the more 3dfx can get developers
to tailor leading edge programs to their chipsets, the more 3dfx
will prosper. Their branding is going well.
The stock price has appreciated nicely, but also consider that
(1) the price had already been knocked down from Sega news and
(2) a lot of the appreciation was deserved when the next gen
3dfx chips became revealed, showing they were still ahead of the
game. The fact that revenue is growing faster than expected is
the real kicker. If 3dfx can get large volumes of their chipset
into the gaming community, it should solidify that market before
the Banshee makes inroads into the general graphics market.
The push to Direct3D and OpenGL will make graphics more
hardware-independent and this is somewhat negative for 3dfx
Glide solutions. But on the other hand, the leading edge of
games will almost always be taping into the latest, greatest
hardware. If 3dfx makes its features relatively easy to program
and their chipsets capture a large % of the gaming market, the
best (funded and quality) games will be optimized for 3dfx to
some degree.
That's all the thoughts for now. Your comments/criticism would
be appreciated.
-Bill |
