Reflecting on past interactions with Samsung Electronics (Samsung) at a book launch event for his new autobiography, TSMC founder Morris Chang noted that former Samsung chairperson Kun-Hee Lee once stated that TSMC would not be able to succeed in memory chips, and should collaborate with Samsung. Chang stressed that he thought collaborating with them was not feasible, and still believes so.
In light of recent developments at Samsung, Chang pointed out that the company faces technical challenges rather than strategic issues. He remarked that the current chaotic situation in South Korea will undoubtedly hinder corporate operations.
In his autobiography, Chang recounted an invitation from Lee for a meeting in Taiwan in January 1989. Lee essentially suggested that the memory industry requires substantial capital and talent, which he doubted Taiwan could provide. To showcase the investments and expertise Samsung had already committed, he invited TSMC personnel to visit Samsung's memory chip factories. If TSMC were to enter the memory market, the best approach would be to partner with Samsung.
In the same chapter, Chang discussed efforts made by WSMC and VIS in the 1990s to penetrate the memory sector—a bold attempt by Taiwan that ultimately failed. Fortunately, TSMC emerged as a savior, preventing significant losses. Today, Samsung continues to dominate over half of the global memory market, and despite Lee's suggestion to collaborate, WSMC did not pursue this path after they failed in memory production. Chang cautioned that the unknown may be more dangerous than the known.
Chang also reflected on his autobiography's account of TSMC's collaboration with Qualcomm several years ago regarding the 28nm process node. At that time, demand for the 28nm technology surged, leading to a considerable gap in supply capabilities. When Qualcomm sought a second supplier against Chang's advice, he responded that by the time its new source could begin supplying, TSMC would already be able to meet all its needs. Qualcomm eventually turned to GlobalFoundries, which offered lower prices.
Chang's prediction proved accurate. There was excess capacity in 28nm, and another client approached him, stating that since cheaper sources have emerged for 28nm chips, they would be reducing orders.
Chang felt disappointed and surprised as TSMC had thought of that client as a loyal partner. Emotions still run high every time he recalls the 2012 experience with 28nm technology. Quoting Shakespeare, he said, "All's well that ends well." The outcome of the 28nm saga marked TSMC's entry into the smartphone world with first-generation technology, and the excess capacity was merely temporary.
