SAN FRANCISCO—Japan's Elpida Memory Inc. Monday (Nov. 1) reported strong year-to-year gains for the second quarter and first six months of its fiscal year, but the company experienced sequential sales decline between its first and second quarters due to eroding DRAM selling prices.
Elpida (Tokyo) reported preliminary sales for the six months ended Sept. 30 of 325 billion yen ($4 billion), up 92 percent compared to the same six months of its fiscal 2009. The company posted a preliminary net income for the period of 39.5 billion yen ($490.3 million), compared to a net loss of 51.6 billion yen in the year ago period.
For the quarter ended Sept. 30, Elpida reported sales of 148.7 billion yen ($1.8 billion), down 16 percent sequentially but up 55 percent compared with the year-ago quarter. The company reported a net income for the quarter of 8.8 billion yen ($109.2 million), down 71 percent from the previous quarter and compared to a net loss of 7.2 billion yen in the year-ago quarter.
Year-to-year first half sales growth was mainly due to an increase in DRAM bit shipments and average selling prices (ASPs), Elpida said. The quarter-to-quarter sales decline from the first quarter to the second quarter was due to declining ASPs and sharp appreciation of the Japanese yen, Elpida said. Quarter-to-quarter bit growth in the second quarter was at the lower end of the company's guidance of 5 to 10 percent, Elpida said.
Elpida plans to post its final results for the first half and second quarter of its fiscal year Thursday.
As long as Samsung builds fabs like there's no tomorrow, and no one else is able to match them, it becomes a question of time before Samsung has a near monopoly. There is one company that is capable of matching Samsung fab-for-fab, or exceed them, but it's too busy buying virus software companies and paying out dividends and doing stock buy backs. If Intel were to do something like merge with Micron, and get into the business of building fabs and becoming the #1 semiconductor company, and not just the #1 microprocessor company, Intel could probably beat Samsung and all other challengers. However, Intel doesn't seem to have the stomach for the fight.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.