When Should You Replace Old Hardware and Software?

13 Nov, 2012 By: Robert Green

The cost of replacing outdated tools is substantial, and management often uses it as a reason to delay upgrades. But you're also paying a price by continuing to rely on those clunkers.

One conversation I have with many CAD managers — more than you’d think — goes like this:

"We run [fill in software name here] version [fill in the year] on seven-year-old computers, because my boss says everything works fine and he doesn’t want to pay for anything newer. How can I convince my boss to update our tools?"

In this edition of the CAD Manager’s Newsletter, I’ll address this all-too-frequent problem from financial, staffing, and marketing perspectives by relating actual client stories and drawing conclusions as we go. Here goes.

Acknowledge the Objections

Because old software or hardware is still in use, senior managers often see no need to replace it — they think, "If it ain't broke, don't fix it." The phrases I’ve heard management teams repeat over and over when defending the continuing use of old software and computers include:

  • The old tool works fine.
  • We’ve always done it the old way.
  • New hardware costs lots of money.
  • New software will require investing training time as well as money.
  • New software requires customization.
  • We’re busy; we don’t have time to implement updates.
  • We don’t want to pay upgrade fees.

My immediate response to all these objections might surprise you: Every one of them is right, to an extent. The thinking CAD manager will acknowledge that management has a point, and that it simply wants to limit spending in any way it can. However, if we never swapped out our old tools, we’d still be using pencils on drafting boards! So it's really not a question of if you should upgrade, but when.

You'll need to answer the inevitable management questions, "How old is too old?" and "How do we know when to change to the newer version?" with some degree of authority. To do so, you'll have to spend some time evaluating your own company's circumstances. You can also use the following real-life examples to illustrate that sometimes, choices that seem like they save money really don't.

Dangers of the Do-Nothing Approach

For the sake of argument, let's say that your company decides to change nothing, and continues to use the old software or hardware forever. What would happen? Here are some consequences that I’ve encountered in the field:

No hardware safety net. I once visited a manufacturing firm that used some mid-1990s CNC software on their shop floor. Upon opening the hardware rack to examine the tool interface, I witnessed a Gateway 2000 486DX/33 computer driving the tool via a 16-bit RS-232 serial interface. Now there are people reading this newsletter who have no clue what I’m talking about, but let me assure you that this computer was at least 19 years old.

My question to the CAD manager was, "What’s your plan for when that machine or serial interface stops working?" and the reply was, "We’ll deal with that when it happens." The CAD manager fully understood the gravity of the problem, but felt powerless to do anything about it because he’d been told "no spending" for years. I told him, point-blank, "When this machine fails, management is going to blame you." At that point we developed a strategy for reporting on the risk of the situation, and senior management began studying the possibility of replacing equipment based on risk management. We also decided that buying a similar vintage machine off eBay for $100 would be good insurance in the short term.

Lesson learned: Hardware will eventually fail, whether you are prepared for it or not, and the CAD manager will take the blame when it happens.

Questions to consider: How much will it cost your company when your antiquated hardware systems fail? Even if you can afford the cost of a rushed replacement, can you afford the project delays and employee downtime that will result from the failure?

Peripheral problems. I have a customer that has modern CAD tools, very progressive management processes, and a well-trained workforce, yet persists in running two 15-year-old plotters in one of its departments. Every time any software or network updates are implemented, these peripherals become a problem, and a mad dash to find updated drivers ensues. Much like the hardware failure scenario above, one has to ask the question, "Why do we depend on outdated devices that cost us time and divert attention away from our projects?"

Lesson learned: Sometimes other system tools receive upgrades while peripherals are ignored — and conflict is always the result.

Questions to consider: What will happen when you can’t plot? Or print?

Operating system failure.
At some point — and you never know when — some sort of system update or service pack may cause your old software to simply stop working. For example, I have to maintain an old computer/compiler for some legacy software I develop for a specific client. For that purpose, I have a dedicated 32-bit Windows machine with old versions of Visual Studio and Autodesk development tools on it. One day, an automatic Windows patch installed itself on that machine — and rendered the compiler environment unbootable! Fortunately, I had prepared a boot disk for the machine and could roll the installation of the patch backward to restore the machine without a complete rebuild.

Lesson learned: At some point, Microsoft no longer cares if your old CAD software still works! Backup disks are worth their weight in gold.

Questions to consider: Are you prepared for operating system problems? Do you have the required backups? Are you sure you can recover?

Indirect IT problems. When your CAD users continue using really old software, you may have to keep them on old 32-bit operating systems, which can compromise your ability to run other software applications. Over the past few years, I’ve encountered everything from AutoCAD R14 on Windows NT 3.51 to AutoCAD 2000 on Windows 98.

In cases where these old operating systems are left in place to support old CAD software, everything from modern e-mail systems to web-enabled tools to SQL database tools are no longer supported. These antiquated machines are now isolated and have to be "worked around" from an IT point of view, which means an inordinate amount of time is spent on older hardware.

Lesson learned: Old software requires old operating systems, which in turn may not support current IT tools.

Questions to consider: If you have to spend hours and hours supporting an old machine every year, are you really saving any money? How much IT money are you spending to keep old machines running?

Customer alienation. When customers start sending you files you can’t open with your existing software, the really serious problems begin. You may be able to use utilities to save down from current versions to your old version, but sooner or later the kinks in your workflow become disruptive at best, and a technical show-stopper at worst. How will you handle it when you get a Revit file from a client and you only have AutoCAD 2006? What about when you get a SolidWorks file from a client and you only have AutoCAD 2008? How do you expect to cope when you get an AutoCAD Civil 3D project, but you only have an old v8 copy of Bentley MicroStation?

If you never share files with outside customers, then this scenario doesn't apply to you, but I see far too many companies that spend way too much time working around these types of problems.

Lesson learned: Even if time stands still at your workplace, it's still moving right along at your clients' companies.

Questions to consider: How can you deliver new work products to clients if you use only old tools? What will it cost you to work around these issues? Do outdated tools make your company seem old-fashioned, less competitive, or even incompetent?

Summing Up

The challenge now falls upon us all to estimate the financial risk our companies are taking when they choose to work with outdated computers, peripherals, operating systems, and CAD tools. By explaining the cost of doing nothing to your management, you may be able to persuade them to start doing something instead.

About the Author: Robert Green

Robert Green

Add comment

Download Cadalyst Magazine Special Edition