Planar 1959 Step 17 Industry Adoption Skepticism Licensing

# Survey Who Adopted the Planar Process in 1959 and 1960, and Who Still Trusted Mesa: The Claim Meets a Skeptical Industry

## 1. Why a Patented Process Still Had to Be Sold to Engineers Who Had Just Finished Trusting Mesa

Step 16 described the claim Hoerni and Fairchild filed; this step describes what the rest of the semiconductor industry actually did in response, because a patent on a better process does not automatically convert every competing manufacturer into a licensee or a convert — some adopted quickly, some negotiated licenses, and some spent real engineering effort trying to reach a comparably reliable device without touching the specific mechanism the patent claimed. Fairchild itself converted its own production lines over 1959 and 1960, replacing the mesa line this project documented in 1958 outright. Other manufacturers who had just spent years tuning mesa etch and coating processes of their own were, initially, reluctant to abandon a known process for one that required a different photolithographic sequence and a different furnace discipline, even with Step 8's reliability data already circulating as a published result.

$$P_{\text{adopt}} = f\!\left(\text{data credibility},\ \text{switching cost},\ \text{licensing terms},\ \text{competitive pressure}\right)$$

where switching cost includes retraining a process-control staff already expert in mesa etch characterization, and competitive pressure rises sharply once one manufacturer's planar-based parts start winning reliability-sensitive contracts the way this series' own Step 8 result predicted they would. The manufacturers surveyed in this project's own mesa-production extension — Texas Instruments, Motorola, Transitron, Philco — each reached a licensing or adoption decision on a different timeline, driven by exactly this mix of factors rather than by the technical merit of the process alone.

Four Manufacturers, Four Different Adoption Clocks time to planar adoption or licensing, relative to fairchild's own 1959 conversion FAIRCHILD immediate, own invention TEXAS INSTRUMENTS licensed, parallel own development MOTOROLA licensed after evaluating mesa alternatives SMALLER MESA SHOPS slowest, held mesa lines longest Padopt = f(data credibility, switching cost, licensing terms, competitive pressure) the same reliability number from step 8 moved each manufacturer on a different schedule

## 2. Real Diagram: The Skepticism Argument, Stated the Way a 1959 Process Engineer Would Have Heard It

A production engineer running a mature mesa line in 1959 was not being irrational by hesitating: switching introduced new photolithographic alignment steps, a new furnace discipline around masking oxide integrity, and a licensing cost, against a published reliability result from a competitor with every incentive to report it favorably. The skepticism eroded only as independent accelerated-life data from licensees themselves began confirming Step 8's own numbers.

Why a Working Mesa Line Was Not an Easy Thing to Abandon the case for staying on mesa a little longer, as a 1959 engineer would have made it mesa process control already mature, staff already trained switching means retraining and requalifying a whole line competitor-reported reliability data, not yet independently confirmed a licensing fee is a real cost against an unproven internal result independent licensee data begins matching step 8's own numbers skepticism erodes, adoption accelerates through 1960 the technical argument this series made was correct from step 1; the industry still had to be shown it independently

## 3. Why Adoption Speed Belongs in This Series, Not Just in a Business History Elsewhere

This project's own mesa-production extension documented the competing manufacturers' mesa efforts in 1958 as a snapshot before this series' process existed to compete with. This step completes that comparison by tracking what each of those same manufacturers actually did once Step 16's patent and Step 8's reliability data were both on the table, which is a different question from how good the mesa process was on its own terms. The variation in adoption speed across Fairchild, Texas Instruments, Motorola, and the smaller mesa shops is itself evidence for a point this series has made since Step 1 in purely technical language: the planar process was not an incremental tweak competitors could absorb casually, it required enough of a process-discipline change that even manufacturers who believed the reliability data still took real time to convert.

Step 17 does not change anything about the process itself; it completes the business picture Step 16 opened, showing that a correct, patented, and published result still had to out-compete institutional inertia on a timeline measured in months, not days.

Take planar 1959 step 17 industry adoption skepticism licensing further

Ask the copilot about this term, or have our engineers assess it against your process.

Ask CFSGPT about this → Request a technical assessment → Open the CFS app