2.3.2. DRAMs

Then we have DRAMs which are about 150 to 100 times slower than SRAMs; that means, to bring a certain amount of data a data unit a word from DRAM the processor will require about hundreds of processor cycles to do so. The speed of a DRAM is typically in the range of 50 to 70 nanoseconds; that is the access time is in the range of 50 to 70 nanoseconds. But, it is also about hundred times cheaper than SDRAMs. So, the typical cost of DRAM units range in between 20 dollars to 75 dollars per GB.

Dynamic Random Access Memory (DRAM) is a type of memory used in computers that is slower than Static Random Access Memory (SRAM). When a processor needs to retrieve data from DRAM, it takes significantly longer—typically between 50 to 70 nanoseconds—often requiring hundreds of processor cycles to fetch a single word of data. Despite being slower, DRAM is much more cost-effective than SRAM, costing around $20 to$75 per gigabyte, making it a preferable choice for larger memory requirements.

Imagine you're at a library (like DRAM), looking for a book (the data). It takes time to walk through the aisles and find the right shelf (equivalent to the slower access time in DRAM). If you were to compare this to finding a book on your phone (like SRAM), it's much quicker to retrieve information online. However, the library has many more books available (larger capacity), and it costs far less to borrow than buying them each individually.

So, to achieve the best performance what would we desire? We would desire a very large capacity memory which can hold all our programs and data and which works at the pace of the processor. That means, if a processor requires a memory word in one cycle it is available in the processor from memory in the next cycle itself. However, in practice we saw the cost and performance parameters and it is difficult to achieve.

In computer system design, achieving optimal performance often involves balancing various factors like memory capacity and access speed. Ideally, we want a memory system that can keep up with the processor speed—delivering the needed data in one cycle. Unfortunately, while larger memory designs can provide space for more data, they also tend to be slower and more expensive, which creates a challenge in finding that perfect middle ground.

Think of it like needing a larger car to seat your entire family (the memory capacity). However, the larger the car, the more fuel it consumes and the more expensive it becomes to maintain (reflecting higher costs and slower performance). In reality, you might opt for a van that's spacious but perhaps a bit sluggish, since it fits your needs without breaking the bank.

So, although the faster memories the mem although SRAMs are very fast in terms of access time, they are also very costly. The solution is to have memory hierarchy where smaller more expensive and faster memories are supplemented by larger, cheaper and slower memories.

To navigate the complexities of cost and performance, computer systems utilize a memory hierarchy. This structure consists of various types of memory organized from fastest and most expensive (like SRAM) at the top to slower and cheaper options (like DRAM and magnetic disks) at the bottom. By using this hierarchical model, systems can balance performance requirements with cost considerations, ensuring that frequently accessed data is quickly available while maintaining overall data storage.

Consider a restaurant that has a different dining experience based on your budget. You could eat at a fancy restaurant where meals are prepared quickly, but it's expensive (fast, costly memory like SRAM). Alternatively, there’s a buffet that’s more affordable, but you have to wait longer for food (slower, cheaper memory like DRAM). This tiered dining system allows customers to choose according to their budget/preferences, just like computer memory does.

--