The Problem and the Alternatives
For quite some time, embedded systems and server designers have relied on a fairly simple memory hierarchy. They've used volatile memory (SRAM and DRAM) for fast data and code storage close to a processor and rotating mechanical memory (hard disk drives or HDDs) for large-capacity, non-volatile storage. As processors got faster, the performance gap grew larger between processor bus cycle times and DRAM cycle times so system designers started to employ fast SRAM as a cache to alleviate some of the access-time problem with DRAM.
There's a similar but even larger performance gap between the access time of DRAM and HDDs. Although vendors have improved HDD capacity by 60% per year—each and every year—and HDD's price per storage bit directly tracks that trend as well, there's been very little improvement in HDD data transfer rate and interface speed and there's been no dramatic change in HDD access time, which is largely determined by mechanical factors. Consequently, there's been only a relatively slow improvement in HDD IOPS (I/O operations per second), which leads to a massive five-orders-of-magnitude (10^5) performance gap between DRAM access times and HDD access times and that performance gap is growing.
At the same time, DRAM's volatility plays a role in a system's sensitivity to power glitches and losses. When data is critical, and most data is critical these days, non-volatile memory just isn't sufficient. Some means of retaining data through a power loss is usually required. In the past, HDDs have sufficed for non-volatile storage but they're simply too slow these days.
The Flash Zone
Flash memory is a good candidate for filling this memory gap because it provides nonvolatile storage and it has become the cost-per-bit leader in semiconductor memory. Consequently, Denali calls this performance gap in the memory hierarchy the "Flash Zone", the performance zone between DRAM and HDD access times.
Multiple technologies have vied to fill the Flash Zone. nvSRAM, such as those offered by Cypress, although a mature and cost-effective solution, are unable to scale to larger densities. More exotic technologies such as PCM, MRAM, and FeRAM also have scaling issues and are not yet mature. Time will tell if these technologies make it out of the lab to high-volume production. Designers have been limited to alternatives such as battery-backed RAM with the associated issues with batteries or living with the issues of slower NV memories, such as Flash, which do not perform well for random accesses or for write transactions. For many applications these are poor alternatives and have caused designers to implement alternative architectures, e.g. caches, to get around many of these issues while still not obtaining the best performance possible. However, now several trends in the industry are making these tradeoffs even more pronounced. Thus there is a growing need for higher density, high-speed non-volatile random access memories. And the traditional approaches are not adequate.
In response to the need for a fundamentally better approach to implementing large, high-speed, NVRAM, AgigA Tech has developed the AGIGARAM® Non-Volatile System (NVS).
The AGIGARAM product family merges NAND Flash, DRAM and an ultracapacitor power source (PowerGEM®) into a highly reliable non-volatile memory system. The innovative memory subsystem meets the needs of high-speed, high-density non-volatile RAM to offer the best fit solution to fill the "Flash Zone".
Learn more about our AGIGARAM technology.