***Extending the Reach of 3D - Points to Ponder
by John Latta

The movement of 3D into the business environment is a high priority. If
demand can be stimulated the 3D industry will move from its current
dependency on the consumer sector and it will be able sell products at much
greater than mass market prices and at the same time avoid the seasonal
market cycles. However, past attempts to drive 3D into the business market
have failed. In this issue we reported on S3's move into supporting desktop
3D and now HP and Microsoft with DirectModel.

The DirectModel effort is part of a larger movement to extend the result of
the design process in a product-oriented company throughout the enterprise.
However, to just provide massive million+ polygon models, even if reduced
by DirectModel, is of little value to the desktop with a lowly Pentium.
Thus, for 3D models to be enterprise pervasive they must follow the same
rules as applications on the desktop: easy to use, integrated with other
applications and not require changing the organizational computing
infrastructure. DirectModel is certainly a move in this direction.

OpenGL Optimizer discussed in this issue also attacks the same problem.
Here we have the typical clash of the market. SGI emphasizes the open
nature of its proposal whose final result will be the action of industry
group it has helped to form while HP and Microsoft have market presence in
both the MCAD world and on the desktop OS (Windows NT), respectively. PTC,
Clarus and SDRC overlap on both groups. HP and Microsoft landed: AutoDesk,
Cadcentre, CoCreate Software Inc., Computervision, EAI, EDS-Unigraphics,
Fraunhofer Institute for Computer Graphics (IGD), Matra Datavision,
Resolution Technologies, SensAble Technologies, Inc., SENSE8 Corporation,
SolidWorks Corp., Spatial Technology Inc., and Template Graphics Software
Inc. Unique to SGI is Dassault Systems and Division.

There are some potentially significant implications in the HP announcement.
Microsoft indicates that it is embracing DirectModel, with its
extensibility, in the content of DirectX. In the past Microsoft has not
advocated open extensibility to its APIs because of the support and driver
issues. There is also an element of lost control by making software openly
extensible. Potentially even more important is the linkage of Direct3D with
DirectModel. Microsoft has not been pushing Direct3D into the professional
space and with DirectModel it implies that it is doing so now.

We see the capabilities of culling and the simplification of complex models
as being a critical move to placing 3D onto the desktop. Significant 3D
processing power will still be required and this is not on desktops today,
even at the high end of the PC space. Going forward we see the 3D pipeline
being extended in hardware to include the extensions now being brought to
market in software - another feature that will keep 3Dlabs and other
companies busy for years to come.

The announcement by HP and Microsoft also shows the reach that DirectX is
having. Originally SGI had licensed QuickTime but this was a flop. Now HP,
not seen as a multimedia heavyweight, has licensed DirectX as a way to
rapidly extend its workstation solutions into multimedia. It will take time
to see how effective the port will be and if it is leveraged by
applications in the workstation space.

All of this indicates how quickly the workstation and Wintel market are
converging. Out of the convergence will be losers and the stage is being
set for marketplace battle.


Wave Issue 9712 5/21/97 Article 7-01