The intent of this analysis and overview of Random Access Memory (RAM) is to define its various types and how they are used in system configurations. Included in this analysis is an overview of the operational characteristics of RAM including how to install and use dual in-line memory modules (DIMMs) and single in-line memory modules (SIMMs). Memory modules of all types also require prescriptive maintenance, and several recommendations are provided in this analysis and overview of how best to keep each form of memory functioning at an optimal level. Troubleshooting tips for managing memory are also included.
Overview of Random Access Memory
The rapid adoption of Random Access Memory (RAM) in electronics products historically has been driven by the exponential increase in integrated circuit design-ins for consumer and commercial products and the compounding effects of Moore's Law (Adee, 2009). Today RAM is designed and manufactured only the form of integrated circuits (Luo, Luo, Guan, Zhou, 2013) with Dynamic Random Access Memory (DRAM) being the most prevalent today. Of the many subsidiary types of DRAM modules the most prevalent is synchronous dynamic random access memory (SDRAM), which is an area that Samsung Electronics continues to invest in heavily as they see this technology being critical to the future of portable devices (Samsung Investor Relations, 2013).
Read Only Memory (ROM) is a very common type of integrated circuit delivered din electronics products predicated on the RAM product design as well (Adee, 2009). Of these two types, DRAMs are experiencing greater use today as consumer electronics products manufacturers increasingly design them in to their circuit broad designs (Raoux, Burr, Breitwisch, et.al., 2008). Best-selling products including the Apple iPod, iPhone, iPad series have lead to DRAM being one of the most preferred technologies for miniaturized, highly portable devices including smartphones and tablet PCs. RAM is especially attractive to product designers in that it is often very high speed in terms of access times (measured in nanoseconds) and manufacturers of these memory modules have developed extensive software and firmware development tools to aid engineering teams (Luo, Luo, Guan, Zhou, 2013). RAM provides many hardware manufacturers with the capacity they need to complete their designs while also delivering high speed access rates, all making the design of state-of-the-art smartphones and tablet PCs achievable.
When the first personal computers were designed, memory manufacturers relied on single in-line memory module (SIMM) configurations of RAM. It was common to see computer motherboards with rows of SIMM modules, with the highest-end workstations and servers having multiple banks of these to support more complex software processing tasks. Dual in-line memory module (DIMM)-based integrated memory circuits replaced SIMMs given their design attributes and characteristics being more aligned with Intel Pentium-class systems and above (Adee, 2009). Today DIMMs support a full 64-bit memory path which is more suited for the high speed bus architectures and microprocessor designs of the current and future generation of computer systems (Luo, Luo, Guan, Zhou, 2013). Today DIMMs are most often produced in the form of dynamic random access memory (DRAM) and synchronous dynamic random access memory (SDRAM). Memory manufacturers have rapidly adopted these standards to support the continual improvements in microprocessor speeds and functions.
Samsung is a technology leader in RAM and DRAM production, also having a the majority of sales in each of these areas. 36% of all Apple DRAM demand is comprised of Samsung integrated memory circuits (Samsung Investor Relations, 2013). As Samsung is a large proportion of the market, their forecasts for DRAM demand worldwide are indicative of broader market growth. Based on an analysis of the Samsung filings with the U.S. Securities and Exchange (SEC) Commission and an analysis of their annual reports, the following DRAM forecast summary has been created in Figure 1. Samsung continues to aggressively pursue cost reductions as shown by the Average Sales Price (ASP) change over time, all of which are double-digit with the exception of 2013.
Figure 1: DRAM Forecast Summary, Worldwide
Source: (Samsung Investor Relations, 2013)
Rapidly Changing Operational Characteristics
Given the rapid price declines that Samsung and other DRAM producers are expecting this and future years, continual innovation is critical. The operational characteristics of RAM are continually changing to reflect the rapidly evolving needs of Original Equipment Manufacturers (OEM) who rely on SDRAM and DRAM for current and future product generations. Current and future operational characteristics include planning and implementing chipsets with low power consumption; support for high speed read, write and cyclical redundancy checking (CRC); support for enhanced wireless technologies including support for a standard wireless Flash interface and support for temperature-compensated refresh (TCR) and partial- and full-array refresh (PAR) technologies (Luo, Luo, Guan, Zhou, 2013). All DRAM and SDRAM integrated circuits are continually being modified to reflect more extended operating temperatures as mobile and tablet PC devices are increasing in overall energy usage. Samsung continues to invest heavily in these areas of extending and enhancing memory usage within operating temperature constraints and ensuring Joint Electron Devices Engineering Council (JEDEC) certification and compliance (Samsung Investor Relations, 2013). All of these operational characteristics are critical to the overall development of DRAM and SDRAM integrated circuits which can be used effectively throughout the design cycles of OEMs including Apple, IBM, Dell and many others (Luo, Luo, Guan, Zhou, 2013). High performance DRAM and SDRAM modules also support integrated packet buffering and look-ahead cache optimization techniques including bus optimization techniques, all of whicha re critical for personal digital devices and the continual rapid evolution of advanced server configurations to support virtualization-based tasks including cloud computing (Samsung Investor Relations, 2013).
Steps To Installing SIMMs and DIMMs
The following steps illustrate how to install SIMMs and DIMMS onto the motherboard of a personal computer. DIMMs have replaced SIMMs given their 64-bit data path relative to the smaller 32- and 16-bit data support on SIMMs (Raoux, Burr, Breitwisch, et.al., 2008).
First, the case of the computer being upgraded needs to be removed. Many technicians will wear a ground strap to ensure that static electricity doesn't accidentally damage the components inside. In production centers there is also an electrostatic mat or carpet that technicians stand on to further ensure they are grounded so static electricity doesn't damage the components. The entire system is unplugged and turned off at this point.
Second, the technician places the system on a secure location so that motherboard can be easily viewed and seen. For larger servers this can involve placing the server on its side to gain access to the motherboard.
Third, the DIMM modules (and on 8088-based systems, these would be SIMM memory modules) are placed into the available slots on the motherboard. Figure 2 shows an example of how to accomplish this. Often the SIMM module slots will be located on motherboard at a location that is close to the power supply.
Figure 2: Installing DIMM Modules On The PC Motherboard
Third, the entire case is closed up and plugged in, booted up, and configured so the systems' BIOS (resident in ROM integrated circuits on the motherboard of the PC) can identify the memory.
Fourth, depending on the operating system, it may also need to be configured to identify and use the DIMM modules. In previous generations of Microsoft operating systems this is often the case.
Fifth, if the system that has been configured is a low-end file or print server on a network, the network operating system, especially in Novell and peer-to-peer configurations, will also need to be modified (Chen, Ranganathan, Pai, Lilja, Bazargan, 2005). These steps ensure that the memory can be shared across multitasking system configurations that rely on a shared memory model, which is common for Novell networks and advanced peer-to-peer networking operating systems (Raoux, Burr, Breitwisch, et.al., 2008).
Sixth, system administrators will also often run a series of diagnostics and memory tests to ensure the system is working at its optimal level. This is especially important in the area virtualized network configurations that are supporting cloud-based architectures (Luo, Luo, Guan, Zhou, 2013). Integrating diagnostics directly into the advanced integrated chipsets used in servers designed for cloud computing environments represents a significant investment opportunity for Samsung Corporation (Samsung Investor Relations, 2013).
Performing Preventative Maintenance And Troubleshooting Memory
Preventative maintenance tasks for memory need to be periodically carried out to keep these integrated circuits functioning at an optimal level. First, many system administrators have purchased aerosol cans of compressed air that can be used to clean motherboards, DIMM and SIMM modules. This works well to get the dust off these chipsets and ensure they will continue to work correctly. Second, measure the level of recommended voltage on the system to ensure each DIMM slot (or on 8088 systems, SIMM modules) have the correct amount of voltage. On older systems the further a SIMM module is from the power supply the less power it receives. For best results the highest capacity SIMM modules need to be by the power supply. Third, check cooling fans frequently…