Nov 19

Shifting Emphasis of Memory Technologies

One of the most interesting occurrences this year was when PCs were reported to no longer consume the majority of DRAMs. That moment happened to coincide with the quarter in which microprocessor manufacturers blamed poor quarterly results on weaker-than-expected PC sales, while the leading cell phone manufacturer reported higher than expected earnings based on a surge in demand.

The angst that was released over the declining role of PCs was a bit surprising. The growth of mobile devices and the corresponding decline of desktop computing devices as the target applications for new memory technologies had been recognized for quite some time. Additionally, the decline of desktop PCs as the major consuming application for DRAMs by no means signals the end of product life for either DRAMs or for PCs.

One positive outcome might hopefully be an end to the perpetual question of how any new and emerging memory technology will be able to survive if it has a higher cost/bit than DRAM or NAND. The declining role of PCs points out that DRAM demand isn’t falling off because the cost per bit is failing to meet the expectations of Moore’s Law. The growth rate in consumption of DRAM bits is slowing because the demand for one of its high-volume applications is declining.

The point to make is that there was clearly a synergistic market relationship established between PCs and DRAMs. When those DRAMs that had been designed for mainframe computers were combined with microprocessors and an efficient operating system, a new OEM product was created that could bring a high level of computing capability to the desktop. It took some very creative and innovative efforts on behalf of the entire infrastructure from the semiconductor industry all the way through the manufacturing and supply chain to make PCs so ubiquitous. The additional DRAM demand for PCs accelerated the associated DRAM manufacturing cost efficiencies that, in turn, drove ever-higher PC demand. The desktop application itself—along with the DRAMs and the various other hardware and software elements—has simply continued to evolve to the point that the application no longer needs to actually reside on the desktop at all.

In a similar evolutionary step, Jim Pappas, who is director of technology initiatives in Intel’s Data Center Group, spoke recently regarding the Storage Networking Industry Association technical working group. “This working group recognizes that [data storage] media will change in the next three to five years. In that time frame, the way we handle storage and memory will have to change… [I]ndustry efforts are under way to remove the bottleneck between the processor and the storage.”

As system-level power consumption has become a higher priority, changing system architectures to replace high-performance DRAM in servers with comparably lower-endurance NAND SSD has opened the door to broader possibilities of memory technologies. There is now clearly a strong effort on the part of OEMs to accelerate major shifts in various system-level architectures. Investors and architectural designers are now vigorously seeking new memory solutions to improve the way we address the storage and retrieval of data—not just cost/bit replacements to existing architectures.

The strong message we believe is being sent today to current and potential investors is that system-level design engineers are aggressively examining new and emerging memory technologies in hope of creating new relationships between the data processing element and the data storage technologies for next-generation high-volume growth opportunities.

We anticipate that the growth of memory technologies—at least for the next decade—will provide new levels of memory performance that enable newer, highly-efficient, data storage tiers.

The message we derive from the declining relevance of PCs is that we have now officially entered the market phase that will accelerate this broader demand for new and emerging memory technologies, see table above.

Bob Merritt is a founding partner of Convergent Semiconductors, LLC. He is a 25 year veteran of semiconductor and communications markets. Starting in the semiconductor industry as a sales engineer for Intel, Mr. Merritt subsequently held senior executive positions in marketing and operations at major semiconductor memory companies including Mostek, Hitachi, Siemens, and Hyundai.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>