A system is not as fast as its fastest component can be, but as fast as its slowest component allows. What would happen if you were told that there is an element that is limiting and that has not evolved for decades, but decades? Well, the bottleneck we are going to talk about is not something of a specific configuration, but rather a common tendency for all computers. Either way, you’re going to find it, and right now it’s insurmountable.
What is the biggest bottleneck on a PC?
The way to measure the performance of one processor against another is to take the same program in both to see how long each takes to run. Obviously, the one that takes the least time will be the fastest. The problem is that measuring performance is difficult today due to the huge number of factors that exist today. We therefore need synthetic benchmarks and the use of applications to test the performance of the hardware and to be able to get a quantitative and qualitative idea of it.
However, there is a bottleneck which is general in all systems as it has been stable all this time. Plus, it’s been walking a tightrope in this regard for years, and every time a new type of RAM and its interface is designed, care is taken to ensure it doesn’t get lost and affect performance. of the central processor and with it is the rest of the system.
Well, if we look at the technical evolution of RAM memory over the past two decades, we will see how:
- Storage has increased to 128 times more.
- Bandwidth is now 20 times greater.
- However, the latency is only 30% lower.
It is the last point that represents the biggest bottleneck so far. Bringing the RAM closer to the processor would be the best option, however, it would make the systems more expensive and although it is also a solution for consumption, it means breaking completely with the traditional way of manufacturing PCs. Although sooner or later we will have to change the way we understand RAM.
Higher bandwidth does not mean lower latency
When a manufacturer gives the bandwidth of a memory, what he is doing is giving it under impossible optimal conditions and therefore of a single continuous transfer in a period of time. However, things are not so simple and it must be taken into account that the memory controller must manage the accesses to the RAM by several cores of the processor and the associated coprocessors.
- The Integrated Memory Controller or IMC has a maximum number of requests it can handle. If in the end it happens that a tolerable amount is exceeded, it slows down, delaying the rest of the requests and creating latency.
- Each new memory access by a different client involves an accumulated latency time.
For example, you will have seen how each new generation of Intel and AMD processors supports faster and faster RAM, but never the fastest on the market. Indeed, it reaches the point where the tolerable latency is exceeded and becomes a performance bottleneck. In addition, overclocked memories have slower communication times, their maximum bandwidth is ideal for certain applications, but they suffer from slight latency problems.
