Acquiring PCs In Today’s Rapidly Changing Market (01jun97)

June 1, 1997

One of the questions we frequently are asked is whether an organization should purchase new PCs now or wait since performance keeps going up and prices are coming down. Obviously, one answer does not fit all situations. The right solution depends upon the age and configuration of existing hardware and the software you are currently using or would like to use.

In general, however, we believe that architects and engineers should be provided with the latest equipment an organization can afford and that older units be moved to people with less demanding tasks. It is somewhat shocking to find that many technical people have better systems at home than they have at work.

Too many financial managers believe that they are saving their companies money when they hold off upgrading systems that have not yet been fully depreciated. It does not take a large improvement in the productivity of an professional who costs $80,000 per year including benefits to justify a new computer that might be an order of magnitude faster than what is currently being used.

The Price/Performance Explosion

The accompanying table tracks the typical PC one could buy for about $3,000 during the past five years. We have used Gateway 2000 for this data since that company has been very consistent in its product offerings. For the past decade or so, the guideline in the computer industry was that price/performance doubled every two years. Welcome to 1997 where it appears that this ratio is doubling annually. A number of factors have caused this change:

• Intel has significantly improved the performance of its microprocessors to the point where they compete effectively with all except the best RISC processors used in UNIX workstations.
• In order to maintain its market dominance, Intel has aggressively priced its new microprocessors. When the Pentium first came out, a 60-MHz version sold for nearly $1,000. Less than four years later, a 233-MHz Pentium II (see related article in this issue covering this new microprocessor and some of its competitors) sells for just $636 in similar quantities. This is nearly an order of magnitude increase in price/performance during this short period of time.
• The huge volume of PCs sold has resulted in incredible economies of scale. While less than 500,000 UNIX workstations will be sold this year, over 80 million PCs will be shipped worldwide.
• Memory costs have plummeted. Until about 18 months ago we were paying about $50 per megabyte for DRAM memory. Supply caught up with demand and prices dropped to under $10 per megabyte. Machines with 64 MB memories are readily available for less than $3,000. In many cases, a large memory can have a more significant impact on performance than microprocessor speed. Likewise, disk storage costs have also dropped like a rock.
• Graphics accelerator cards have come down in price to the point where the PC manufacturers include them in the base configuration.

Getting The Most For Your Dollar

Having spent a fair amount of time watching the PC evolve, we have a few suggestions on procuring new PCs.

• You are better off buying a system somewhat below the top-of-line and then replacing it more frequently than you would if you bought a more expensive system.
• Install plenty of memory. Most CAD and GIS packages thrive on memory and at $8 to $10 per MB, you can easily afford 64MB or even 128MB.
• Do not ignore a high performance backup device. With 3.2 GB disks commonly used today, we recommend a 4 GB cartridge tape drive.
• Buy the entire system from one source. Your time is too valuable to be spent configuring hardware and software for your computer to work.

It’s Show Time! (01may97)

May 1, 1997

The A/E/C Systems conference and exhibition is this industry’s showcase event for the technologies, products, and personalities that are shaping it today and propelling it toward the future. Historically, the A/E/C Systems show has been the launch venue for a number of very prominent products over the years, and this year will probably prove this true once again.

We are looking forward to attending A/E/C Systems this month in Philadelphia scheduled for June 16-19. This year’s event is expected to draw more than 25,000 attendees and between 400 and 500 exhibitors. This show can be a grueling ordeal, just based on the sheer number of vendors and products, but it does let us see most of the major AEC and GIS players under one roof for objectively evaluating their offerings.

We strongly encourage you to also consider attending so you can evaluate the products and question the vendors as an informed potential buyer, because one product does not a comparison make. To really get a good feel for what different products can really do, you have to see them operate with the same data, constraints, or parameters (all preferably yours) side by side on a level playing field.

Beware, though, that just about any vendor worth its salt can make its product look good -- especially if it’s the only one being evaluated and "verbally compared." Even in this verbal comparison, you must be ready to spar with the tough questions. Try to be defensive and objective, because many vendors will tend toward being offensive and subjective about their products, especially when pressed how their products stack up against the competition.

As much as we would like to think otherwise, and as much as vendors would like you to think otherwise, all software products, AEC or otherwise, are not created equally. Since this is a market that is in a seemingly constant state of flux, many customers often approach vendors with a certain air of skepticism and uncertainty. Unfortunately, these feelings all too often play into the hands of unscrupulous sales people who try to assure potential customers that only they have the solution that the customer before them is seeking. In effect, this all too often opens the door to snake oil marketing ploys by vendors. Of course, it’s always the "other" vendor’s direct sales people and VARs who are at fault as the snake oil purveyors. Surprisingly, though, it is often those finding fault with other vendors that fail to support the claims they are making for their own products.

So, then why does the best product often have few takers? Sadly, in this industry, as in virtually all others, it’s often not the content of the product that sells, but the packaging. Once again, it is not a case of who has the best product, but rather who does the greatest volume and better job of splashy, memorable marketing that comes out the winner in the product sales game.

Jeffrey Rowe, Editor

Is Something Wrong with the CAD Industry? (01mar97)

March 1, 1997

The overall computer industry is booming around the world and the overall CAD industry grew between 15% and 20% in 1996 depending upon what you include. Technology Automation Services, the parent organization to A-E-C Automation Newsletter, tracks the financial results of public-traded companies in the overall CAD industry. For the quarter ending December 31, 1996, the 32 companies in its database reported that revenues increased an average of 24.4%. The average would have been substantially higher except for three companies that are primarily AEC-oriented: Softdesk, Eagle Point and Intergraph.

Softdesk reported quarterly revenues of $8.0 million, down 30% from the same quarter a year ago and a loss of $1.7 million. Eagle Point had revenues of $4.1 million, down 18%, and a loss of $0.2 million. Intergraph had revenues of $294.1 million, down 2%, and a loss of $33.6 million. Just as we were going to press, Autodesk reported its results for the quarter ending January 31, 1997. Revenue was down 9% to $115.0 million and while the company earned $6.0 million, that was a little more than a third of what it earned in the same quarter last year.

If the rest of the industry is doing so well, what’s wrong with the AEC sector? According to Cambridge, MA-based Daratech, overall AEC revenue was basically flat in 1996. Not everyone is hurting, however. The problem we have reporting data on the industry is that the healthier companies tend to be privately held. As an example, Bentley Systems claims that software revenue grew 37% in 1996 and is now running at a better than $120 million annual rate. GEOPAK, a fast growing vendor of civil engineering software stated that it grew more than 40% last year and Cambashi Limited, based in Cambridge, England recently prepared a report that shows the plant design software market grew 12% in 1996 with nice increases at CADCentre, EA Systems and Rebis.

We believe that there are several reasons for the overall slow growth in this market:

• Residual problems associated with AutoCAD Release 13 and the long cycle between release 13 and Release 14. Since Autodesk obtains most of its revenues from new licenses and upgrades, without a new release in over two years, revenue from this source has dropped off. This has a related impact on Softdesk and Eagle Point since their update revenue is closely tied to that of Autodesk. In addition, low-cost alternatives such as diagramming software from Visio are eating into the traditional CAD business Autodesk and its third-party developers have historically enjoyed.
• The lack of exciting new AEC technology. About eight years ago, the mechanical CAD industry began a major technology shift to feature-based solids modeling. It took awhile for the technology to catch on, but now it is the accepted way to design everything from bicycles to jet aircraft and companies such as Parametric Technology, Dassault, SolidWorks and Structural Dynamics are booming. The comparable impact on the AEC market will come from a new generation of object-oriented design tools. Bentley’s Objective MicroStation and Autodesk’s ARX will eventually lead to an entire new class of applications and should result in substantial growth in the 1998-2000 time period.
• The difficulty of justifying advanced CAD technology. This newsletter has been making the point for years that users cannot just look at CAD technology as a way of reducing design costs. That can be done with some fairly inexpensive tools. Instead, both design firms and operators need to understand where intense use of the latest technology can result in facilities that can be constructed at lower cost, operated more profitably and updated more easily. The benefits of squeezing 5% more production out of a petrochemical plant significantly outweighs the savings associated with reducing design costs by 20%to 30%.

The bottom line is that some companies are struggling, others are doing well but that the overall AEC industry has some hurdles to overcome before growth can get back on track and approaches its inherent potential.