Intel’s departure from the NAND flash business is somewhat reminiscent of its exit from the DRAM market in the 1980s.
Intel is at memory business crossroads, again. According to The Wall Street Journal, the semiconductor-industry giant is selling its NAND flash unit to SK Hynix for $9 billion. Besides NAND flash chips, the deal encompasses Intel’s solid-state drive business and wafer operations, including Intel’s fabrication facility in Dalian, China.
The deal includes Intel’s fabrication facility in Dalian, China.
Intel’s love–hate relationship with memory technology is the stuff of legends. In 1968, two Fairchild alumni Robert Noyce and Gordon Moore, founded Intel to focus on silicon-gate MOS memory chips and multi-chip memory modules. At a time when almost every semiconductor outfit was focusing on logic devices, Intel saw how memory storage was stuck in the pre-transistor era.
Then came the birth of DRAM in 1969. More than a decade later, after the onslaught of DRAM chips from Japanese suppliers, Intel COO Andy Grove asked CEO Gordon Moore, “If we got kicked out and the board brought a new CEO, what do you think he would do?” He would get us out of memories, Moore answered.
That day, in 1985, the two men decided to take Intel out of the DRAM business, writes Michael Malone in his book The Intel Trinity. That historic conversation was happening when Intel was at the cusp of probably the biggest success story in the technology business: microprocessor sockets in the IBM PC.
Still, Intel went on to pioneer another groundbreaking memory technology: NOR flash. The company launched the first NOR flash chip in 1988, which subsequently began replacing the EPROM products. However, while Intel started making strides in NOR flash, frequently used to hold software code or for bootstrap memory, NAND flash began its dramatic rise to contain data in consumer devices like the iPod.
In January 2005, IC Insights forecasted that the NAND flash market would overtake NOR flash amid the former’s growing use in MP3 players, USB drives, and other storage devices. Intel saw the writing on the wall and cobbled a NAND flash partnership with Micron.
Why Micron? While Micron wanted to move beyond its dependence on the DRAM business, it didn’t suit Intel to build or buy new fabs to make NAND chips. On the other hand, Micron could build fab shells in its memory chip fabrication facilities. IM Flash, the partnership between Intel and Micron announced in 2005, has been part of the memory industry’s relentless drive to create ultra-dense NAND flash chips.
Intel surpassed many memory capacity thresholds in collaboration with Micron. That includes 3D NAND technology, which stacks flash cells vertically in 32 layers to achieve 256 Gb multilevel cell (MLC) and 384 Gb triple-level cell (TLC) die in a standard package. Micron acquired IM Flash in January 2019, which in some ways, was a harbinger of things to come.
EE Times recently explored the arc of memory technology in its Special Project: The Memory Market.
In the end, Intel, which is now divesting its non-core businesses, decided to move away from the flash memory market known for disruptive pricing and aggressive strategies. However, the company isn’t done with the memory technology yet.
Intel plans to keep the Optane memory technology that it’s been developing in partnership with Micron. The 3D XPoint technology—which Intel markets under the brand name Optane—promises to transfer data faster than NAND flash.
However, for the memory market at large, Samsung and SK Hynix can now play the DRAM and flash game Korean style.
This article was originally published on EDN.
Majeed Ahmad, Editor-in-Chief of EDN, has covered the electronics design industry for more than two decades.