Unfortunately, legacy issues more than optimizations often drive the selection of design tools — prior experience, tool familiarity, established design flows, and vendor name recognition. Also unfortunately, design tools have long been considered a stand-alone capital asset, independently purchased by engineering, and administered and maintained outside of the company's IT infrastructure. Classic IT metrics such as return on investment (ROI), total cost of ownership (TCO), and price/performance tradeoff have rarely been applied to the purchase and deployment of EDA CAD tools.

That situation is changing. Although engineering productivity still sits at the center of tool purchasing decisions, evaluating the true costs of a design environment now presents a more complex problem than simply looking at the price of the individual tools. IT-type metrics are becoming increasingly important in the analysis of CAD tools as companies move to quantify the costs and benefits of make-versus-buy, in-housing, off-shoring, outsourcing, and the purchasing of intellectual property (IP).

The cost of doing business

Engineering tools are not an end unto themselves. Rather, they serve as a component — albeit a critical one — of a business flow that begins with investments in a product concept, and ends with the sale of a completed product. As such, CAD tools should be subject to the same business analysis and continuous process improvements commonplace in every other aspect of business management.

What are the goals of this process improvement? There are six generic objectives:

• To achieve a measurable (and hopefully dramatic) ROI.
• To rapidly recoup tool investment.
• Speedy revenue growth in combination with reduced operating expenses.
• Increased flexibility to changing business conditions.
• Operational efficiency in a global economy.
• Minimizing risk.
• Improved working conditions (“Quality of Life”).

Companies that minimize risk while constantly seeking to improve working conditions are likely to retain their single most valuable investment: personnel. Successful management knows that personnel stability is a crucial prerequisite for continuous product improvement, so staffing concerns need to be part of the business analysis when thinking about the costs of CAD tools.

Return on investment

As mentioned, two metrics have long dominated the financial analysis of end-user software investment: ROI and TCO. These acronyms are often interpreted intuitively, but a more rigorous treatment can be extremely instructive.

In principle, calculating ROI is simple:

Earnings and investment are incremental quantities: “If I invest X dollars in a new system, what additional profits will I obtain?”

Of course, the time period over which these benefits accrue is critical, particularly first-year ROI. Limiting the payback period to the first year allows the resulting productivity increase to set the baseline for subsequent periods. Meanwhile, there is a more sophisticated expression for ROI:

Where:

A = capital costs of software, hardware, etc.
L = monthly labor cost
C = training time in months (considered lost productivity)
D = productivity loss during training
E = productivity increase with new system

This more elaborate ROI equation incorporates the loss of productivity as users ramp up on a new system, along with the costs of purchasing the new system, the costs of training and, of course, the enhanced productivity gained from upgraded tools. This calculation and analysis provides a more detailed look at the true return on investments in new CAD software. (see Figure 1)

Figure 1 — True return on investment on CAD software

Although ROI calculations are essential, they often blur the more fundamental question: “How do I do more with less?”

Resources are limited, both in terms of financial resources and human resources. It's this kind of "clean-sheet" thinking that has led to the rise of outsourcing and, more recently, to offshoring. Budget considerations for these strategies, versus supporting new tools and skills in-house, must evaluate the trade-offs between tool costs and engineering head counts. The analysis is more complicated than a simple process improvement analysis.

Outsourcing and offshoring both suggest that labor costs are no longer directly related to productivity — labor costs vary tremendously from one labor market to another. These variances must be further adjusted, therefore, for the different capabilities and different management and interaction costs across geographies.

Total cost of doing business

Since the mid-1980's, the concept of total cost of ownership (TCO) has been widely promoted by Bill Kirwin of Gartner Group (San Jose, CA). Unlike the differential earnings model suggested by ROI calculations, TCO considerations attempt to add up the direct and indirect costs incurred throughout the entire life cycle of an asset.

Those costs may include purchasing, deployment, training, operations, support and, eventually, retirement of the tool. The costs also may include direct hardware and software costs, operations costs (including CAD managers), administration costs (including user training), and the indirect costs of lost productivity due to unexpected downtime and other computer-related distractions. TCO calculations help illuminate the multiple, hidden costs of owning design tools.

"For a lot of companies today, the only way to stay competitive is to keep taking advantage of the same technology that their competitors are using, as new products continue to hit the market. And some organizations may not even want to know how much that's costing them," says James Delmonte, president of JDA Professional Services Inc. (Houston, TX), an IT staffing firm that also performs cost-benefit consulting.

A TCO analysis improves an engineering team’s ability to execute on the strategic business goals set down by management. However, the success or failure of the company—though frequently a consequence of the success or failure of the engineering team—is the responsibility of the senior management team.

Senior management may consult with CAD and engineering managers in the process of making decisions with financial implications, but issues such as headcount and budget tradeoffs extend beyond the responsibility and vantage point of group managers.

The impact of globalization

With the continuously decreasing barriers to communication, travel, education, and capital investment, the IC design industry is poised for the explosive growth of offshore engineering. This trend is the natural continuation of the globalization of semiconductor manufacture, following the example set in recent years by the offshoring of IC mask layout.

Companies wishing to take advantage of the cost benefits available from globally distributed product development teams face a number of challenges. Some of these — communication, management, and staffing — are universal and well understood.

Other challenges, such as the selection, rollout, and management of EDA design tools, are unique to the IC design industry. Clearly, there are additional requirements placed on design tools able to successfully support an international, distributed user community.

The requirements are diverse: internationalization and localization of design tools and computer platform; quality of support, training, and documentation; licensing and other constraints on ownership; integration into geographically diverse workflows, issues of access control (security), and the robustness and reliability of wide area and virtual private networks (WANs and VPNs), to name just a few.

In addition, it's becoming increasingly obvious that distributed design teams require face-to-face interaction at regular intervals in order to preserve a sense of team and common purpose. The concepts of globalization are intriguing and increasingly a reality, yet the human side of the globalization equation remain unchanged.

Companies committed to a distributed design effort, either through outsourcing or offshoring, need to pay attention to the logistics of distributed tools and project partitioning. But they also must attend to the realities of differing time zones and human interaction across cultures and language.

Shopping for EDA tools

Having laid out a variety of criteria, financial and otherwise, to keep track of while assembling a design environment, it's appropriate now to consider the list of features that deserve significant attention while purchasing the tools themselves.

For starters, if you don’t have a reliable, usable computer platform, no design tool will get the job done. Furthermore, having to rely on consultants to configure, install and maintain application-specific platform configurations is both expensive and a migration risk for the future. Take great care in choosing the computer environment before purchasing the tools.

The second crucial consideration must be the ease of use of the tools. Tools that maximize operating efficiency, while offering reduced training costs, are ideal. Localized, multi-language user interfaces should be standard with the tool to allow comfortable use on internationalized platforms, while ease of deployment, administration, and maintenance are also important.

Additionally, another aspect of a tool often overlooked involves the integration of the tool with other workflow applications including e-mail, documentation flows, backup, and disaster recovery. Look to these details beforehand to reduce frustrations after the purchase is complete. Then buy tools that are easy to use, even if only used on occasion—tools that when revisited are like a bicycle. You never forget how to use them.

Then there is the issue of reliability. Just as the computer platform must be reliable, so the tool application itself should be dependable as well. If a tool crashes or is otherwise unavailable for use, no amount of built-in features or customer support will make up for lost time to get the design completed accurately and on schedule.

Customer support should, nonetheless, be a key consideration in the purchasing decision. The cost, quality and responsiveness of the vendor’s customer support — including availability and quality of documentation — are important factors.

Tool robustness and intuitive usage play an indirect role, by minimizing support needs. Support should be considered a form of insurance: You hope you don’t need customer support, but having reliable support options will mitigate the risk of catastrophe.

Next comes a host of questions related to the flexibility of ownership. How is the tool sold? Are licenses perpetual or time-bombed? What licensing options exist — node-locked, multi-CPU, network, or commuter — and how flexibly can policy restrictions be imposed?

What restrictions exist on WAN/VPN sharing of licensing, across countries and time zones? Knowing the answers to these questions before making the purchase will ease the worries and woes for everyone involved after the purchase and implementation of the tool.

If the design tools will be utilized by a team of more than one, features within the tool that support group workflow may be decisive with regards to the success of the project. Those features include revisions control, access control, and policy controls for IP management.

Also, as workgroups are commonly separated geographically, support for WAN data exchange, and robustness in the face of unreliable network connectivity, are now critical. Before purchasing the tool, find out what will happen if network access to the license server is interrupted. Know in advance if work will be lost in those circumstances.

Standards are always important when it comes to design tools and environments. Using tools built to industry standards, which are non-proprietary, increases the ability to incorporate those tools into a diverse array of new or existing flows and minimizes the risk of “data lock-in” in the future.

Is there an available, local pool of engineering talent that already has familiarity with the tool? Although this can appear to be a two-edged sword — after all, potential competitors value the same skill set that you do — in the long run, a pre-trained labor pool allows more flexibility in pursuing new projects and provides a more stable environment to complete them.

Inevitably, inexperienced staff facing a new tool or experienced staff facing a feature upgrade will require updated skill sets. The fastest and lowest risk approach is vendor-approved training. Find out what training options are available to you, and if they will occur on-site or off-site. If there are tutorials, training programs that can be customized to your site, or academic courses available in the area—know in advance of buying the tool. Don't purchase the tool and then explore the training option.

Naturally, the capabilities of the tools cannot be ignored. However, unless all of the previously discussed issues are satisfactorily addressed, even the most intrinsically capable tool will go to waste. It's essential in the tool evaluation process to have a clear set of evaluation criteria; ideally these are developed before any evaluation actually begins.

Ironically, one frequently overlooked feature is the speed of the tool. Engineers or their managers often start the tool search process with a laundry list of features they believe they need. But that list rarely includes the ergonomics or speed with which the sought after features are actually executed by the tool. This is a common mistake.

Finally, there is the qualitative issue of vendor reputation and history. The purchasing decision should include an evaluation of the vendor's business practices, maturity of the product portfolio, commitment to further development, frequency of versions and updates, rate of incorporating new features, and vendor receptiveness to customer-driven improvement. Most importantly, find out how quickly defects in the tool are repaired and new service packs released (see Figure 2).

Figure 2 — Concerns vary among different groups

Be not afraid

Choosing the right CAD tools is extremely important to the success of each and every design project. While most engineering managers can readily identify the technical requirements of the tools needed, other requirements — financial, platform, and ergonomic — are often overlooked when an opportunity for making a collective purchasing decision is missed.

The long-term solution for success in assembling a viable design environment must involve a triangle of decision makers — senior management, CAD managers, and engineers — each offering candid input and a willingness to compromise, so that the constraints of all of the stakeholders involved are met to the best ability of the combined group.

At Tanner EDA, we are often called in to mitigate unfortunate situations where customers have gone into a project ill prepared both financially and technically. We see over and over again that these problems could have been avoided, if appropriate and methodical decision making processes had been applied at the outset — not just with regards to the scope of the project, but more fundamentally with regards to the tools and the resources available to support the project.

The criteria enumerated in this article should help all design teams identify and optimize to their particular constraints, and plan for success. Assembling a design environment is not rocket science — it's common sense.

Massimo Sivilotti, Chief Scientist, Tanner Research, is responsible for setting strategic directions for Tanner’s EDA tool development, developing specifications for new feature development, and working with customers to support their evolving needs. Sivilotti has been involved in microelectronics design and EDA tool development for more than 20 years.