Compare and contrast RAM types and features

 Compare and contrast RAM types and features


1.3 Compare and contrast RAM types and features

Memory is one of the most important parts of a PC but differs a bit, depending on the board.

    1. DDR (Double Data Rate)is a type of SDRAM in which information is sent on both the rise and fall of the clock signal. It is 2x the speed of SDRAM (double pumping). It operates at 2.5 volts and 184 pinsddr

    1. DDR2 iseven faster than DDR (2x). They have 240 pin DIMM (Dual In-line Memory Module)ddr2

    1. DDR3 istwice as fast as the DDR2, therefore 4x faster than DDR .Italso has 240 pin DIMMDDR3

    1. SDRAM (Synchronized Dynamic Random Access Memory)has 168 pin DIMM and operates with 3.3 volts. Its run at clock speed, meaning it synchronizes with the bus speedSDRAM

  1. SODIMM (Small Outline Memory Modules) is an alternative to a DIMM,which is significantly smaller (68mm x 32mm) . Theyare utilized in notebook printers and routers because of the space restriction.SODIMM
      1.  72-pins – 32-bit (no longer in use)
      2.  100-pins – 32-bit, two notches
      3.  144-pins – 64-bit, 1 notch centered
      4. 200-pins – 64-bit, 1 notch  off center
      5. 204-pins – 64-bitt5, 1 notch off center


RAMBUS

rambus

  1. Rambus In-line Memory Module (RIMM) slots came out in the1990s but are no longer being used due to the high expense
  2. Rambus Dynamic Random Access Memory (RDRAM) modules are DDR memory, about 133mm x 35mm in size with a 16-bit, 184 pin or 32-bit, 232 pin form factor

PCS that are compatible with RIMM need to have all of the memory slots filled. RIMM modules, or a Continuity RIMM (16-bit systems), or a Continuity and Termination RIMM (32-bit system) should be placed in empty slots to ensurea continuous signal.

The main difference between SIMMs and DIMMs is that SIMMS electrical ncontacts are connected while DIMMs are on either side of the module. DIMMs are usually 64-bit and come in 3 different pin configurations:

Parity vs. non-parity

Parity is defined as being even or odd. In computer terminology, a parity check usually refers to the method of examining information for errors by setting a parity bit. If a stream has an even amount of “1’s” and the parity is odd, the PC interprets it as corrupt and produces an error.

Parity memory uses an extra chip (to calculate parity) but has a higher cost because of this. It will detect errors but is unable to correct them. Non-parity, on the other hand, has less chips but it doesn’t need to calculate parity. The major downside to non-parity is its intolerance to errors, meaning that the computer will often freeze. Manufacturers have the option between parity and non-parity memory, but no longer use them because of the high cost and unreliability.

ECC vs. non-ECC

ECC (Error Correcting Code) PCs are able to detect errors and correct them. They are usually used in higher-end PCs, as not all motherboards are compatible. Non-ECCs are not able to correct errors but are“cheap”.

ECC memory is a type of computer information storage that can detect and correct the most common data corruption errors. ECC memory is now being utilized on almost every PC where corruption cannot be accepted under any situations, such as scientific or financial computing.

The result of a memory error is system-dependent. In systems without error-correcting code, an error can lead to the “blue screen of death”. Memory errors can cause safety weaknesses.

A single-bit error could be what stops a machine from doing a parity check – this could be ignored in a system without ECC. Immunity-aware programming, Random Access Memory Parity memory and ECC memory are some approaches used to deal with unintended bit flips.

This problem can be mitigated by using DRAM modules that have additional memory bits and memory controllers that utilize these bits. These additional bits are used to take note of parity or to use an error-correcting code (ECC). Parity permits the exposure of all single-bit errors. The most popular error-correcting code(SECDED Hamming code) allows a single-bit error to be corrected and double-bit errors to be detected. Chip kill ECC is, however, better because it also corrects multiple-bit errors, including the loss of an entire memory chip.

In some cases, systems with a non-ECC memory controller can achieve lot of the benefits of ECC memory by using EOS memory modules. An ECC-capable memory controller, as used today,can detect and correct errors of a single bit per 64-bit “word” and identify

(but not correct) errors of 2 bits per 64-bit word. In some PCs, the BIOS, when matched with OS, allows for the counting of detected and corrected memory errors.

Error detection and correction is contingent on an anticipation of the errors that can occur. We have presumed that the failure of each bit in a word of memory is independent and two instantaneous errors are not likely to occur.

DRAM memory can give higher protection from soft errors by relying on ECC. Such error-correcting memory (EDAC-protected memory) is predominantly appropriate for high fault-tolerant applications, such as servers. Some PC salso wipe the memory – occasionally reading all addresses and rewriting corrected versions, if necessary, to remove soft errors.

There are advantages and disadvantages to protecting yourself from loss of information, and the high cost associated with it.

ECC safeguards against concealed data corruption, and is used in PCs where corruption is intolerable, as with most applications in finance and scientific research. ECC also decreases the number of crashes that occur, mainly undesirable in multi-user server apps and maximum-availability systems.

RAM configurations

Single channel- can only access one module at a time

Multi-channel memory architecture increases the information transfer rate between the DRAM and the memory controller, by increasing the number of channels available to communicate.This increases the information transfer rate by precisely the number of channels available. Dual-channel memory uses two, so therefore increases it 2x.

Dual-channel RAM needs a board with dual-channel technology. It also needs minimum of 2 memory modules. It’s best to use memory modules that are identical. In dual-channels, memory banks are normally colour-coded.

When operating in triple-channel, memory latency decreases due to interleaving. Thatmeans each module is accessed in sequence for small bits of information, rather than completely filling up one module before reading another one. Information is spread amongst the modules in an interchanging pattern, possibly increasing memory bandwidth 3x.

The term single-sided is used to describe a single rank of chips that the PC can access at once. While double-sided has chips divided into two ranks, however, ranks are accessed one at time, so it would have to be constantly switching access from one to the other.

RAM compatibility and speed

When it comes to RAM, there are things which are compatible and then things which should never be mixed . First, ensure each stick has matching CAS latency, timings, and voltage. Even though you can change the settings in the BIOS to make the two sticks match, it is not advisable because of over-clocking risk.But, if you are going to try this, you will most likely have to under-clock one of your DIMMS.

Mixing RAM speed is a bit different, however. If you’re mixing speeds, it ensures that they have the same CAS latency, voltage and timing. To cope with the speed difference, your board would most likely under-clock the faster one. Note that the RAM will always choose to under-clock the faster DIMM, unless you manually over-clock the slower DIMMs. Even though this is possible, there is no guarantee – you could do everything right and still be greeted with the “blue screen of death”.

Read more

Practice Test

 Computer Hardware Technician Interview Questions

Sample Resume