How to re-FORM or refill a fab

Fill 'er Up To Make Money

I usually try to ignore items that are unattributed, however a recent blog posting in the ElectroIQ blog “How To Fix FORM” caught my attention. It is true that FormFactor’s current difficulties are being discussed widely. However, the simplistic analysis and suggestions of this unknown “industry insider” need a reality check. The writer gets some of the overall problems right but may be missing the boat on the solutions.

Here are the supposed anonymous industry insider’s suggested fixes:

  • Consolidate manufacturing at Livermore and close Korea and Singapore manufacturing. This is clearly underway. In fact, FormFactor announced on September 13th – a full 15 days before the ElectroIQ blog posting – that they were stopping the manufacturing transfer to Singapore and these operations will remain in Livermore.
  • Define core customers and Tier 2 customers – and focus all efforts on the former. I too believe it’s always critical to prioritize your customers and focus your efforts. However, if there is profit to be had in servicing your non tier 1 customers you shouldn’t ignore them especially in light of my comments below on fab utilization.
  • Tend to knitting until the next upcycle (1-3 years). It is not clear what is actually being suggested here. Substantial R&D, including innovation and product development activities, is required by the continuous need for improved capabilities and new products often expressed as the vicious “technology treadmill” created by Moore’s Law. Without this, the probability of remaining a viable business in 1 to 3 years is low. And a potential acquirer (the second part of this suggested fix) with the intent to rebuild the business would be foolish to do so without a substantial product pipeline. Of course, there are some who might acquire FormFactor simply for the cash on hand.

Beyond these concerns about the suggested fixes, it is not clear to me that this industry insider really appreciates the complex issues involved in running a fab due to his suggestion to “get down to a ~$40M B/E Q run rate”. This reduction of almost 50% in expenses is a lot easier for an outsider to propose – than for FormFactor or any other company in the same situation to accomplish.

It is a no-brainer to reduce variable costs and moth ball start-up offshore manufacturing operations. Unfortunately, this is usually a small part of the overall operating costs. So, why are fab owners, like FormFactor, unable to rapidly cut their costs further? The crux of the problem is that fabs have a relatively rigid cost to operate based on the required capability and designed capacity. Fabs have both fixed costs (building, equipment, infrastructure, etc.) and variable costs (operators, maintenance, supplies, material, etc.). But many of the variable costs are tied to a minimum run rate. Therefore they are actually closer to a fixed cost since without them the fab is non-functional.

For example, certain chemicals for photo lithography and plating have a fixed life. Once opened, they may only last a few hours after which they need to be replaced. At low volume, time alone determines consumption of these chemicals and not volume. And the rate of consumption is made worse by capability – a 300 mm fab will use more chemicals than a 100 mm fab with the exact same processes – since more chemicals are required by either the wafer area or the size of the equipment and “baths” (tanks in which the processing takes place). All of these factors contribute to the minimum operating cost of the fab.

In addition, due to the wide range of processes and equipment that need to be operated and maintained, there is also a core set of operators, technicians, maintenance personnel, and engineers required to operate the fab. A photolithography engineer may not have the right skills to also support the electrochemistry needs of the plating area. Some semiconductor equipment also needs to be maintained on a given schedule independent of volume which also drives the minimum head count for the fab.

Ideally a fab needs to be utilized at high 80% to low 90% rates to be efficient – both in terms of operation and finances. When fab utilization drops significantly, profitability drops rapidly. Therefore, FormFactor’s fundamental problem appears to be fab utilization. If so, like anyone else with this problem, they need to re-engineer their fab to reduce their burn rate and/or figure out what else they can build to consume their fab capacity. At the same time, they need to increase their sales unit volumes (at proper margins) to increase their fab utilization while they continue product development. This is why ignoring non-Tier 1 customers is a mistake since these incremental sales will drive volume in the fab. None of these are easy tasks, so focused management attention for proper execution is essential.

FormFactor and many of their competitors are going through difficult times and need quality advice not Monday morning quarterbacking from an anonymous “industry insider”. I look forward to reading FormFactor’s latest quarter’s results on October 26, 2010. Hopefully they will show that management is taking the company in the right direction.

2 thoughts on “How to re-FORM or refill a fab”

  1. Great points Ira. When the problem is high fixed cost there are two obvious remedies: increase absorption or eliminate the fixed cost. Since the fab is core to their current business model, eliminating the fixed cost does not appear to be an option for Form. Their challenge then is to find a way to increase utilization. Ignoring customers doesn’t sound like a good strategy for accomplishing this.

    In fact, I think your point about developing new products and improving existing ones is key. Maybe they should even look at completely new MEMS appplications or revising their business model? Unless the “knitting” he refers to involves some major recontruction I doubt it will serve to meet their needs.

    There is one other observation I want to make about the “insider’s” original post. He mentions that one option is to get Form “Lean enough to sell”. I fear this comment reflects at best a general misunderstanding of what it means to be “Lean”, and more likely a horrible misuse of the word.

    Getting “Lean” simply means eliminating anything that does not add value. And all companies should always be seeking to do this all of the time, not just before a sale. But in this case, as you point out so well, the ability to innovate and commercialize new products is an inherent part of the business value add equation. I fear, however, that too many investors and corporate lingo slingers think that getting “Lean” means removing some of the business value add capability. If this is true, then their version of getting “Lean” is actually a value reduction activity, and so is the LAST thing one should do before a sale, and has absolutely nothing to do with “Lean”.

  2. Somehow I missed this feedback to our blogs, my apologies…

    I checked back with that “insider,” who simplified/reiterated his belief: FORM’s available market isn’t big enough to really drive profitability, and he doubts there’ll be a friendly technology shift ahead for them to retake market share in probe cards. Assuming this insider’s pessimism about TAM, FORM would need to either take share or cut costs…and given recent trends (as you recently noted) the former seems unlikely. …Though to your argument, he didn’t really offer specifics on how to achieve those goals beyond some ballpark numbers, and he admitted those may not be achievable…

Comments are closed.

%d bloggers like this: