The use of cache stems from the fact that your main system RAM runs much slower than the *www.tomshardware.com/forum/252578-28-what-cache#core. If the CPU had to wait for the slow main RAM for every new byte of instructions or data, it would be ridiculously slow. Ideally, you would have all your main RAM on the CPU chip itself, running at the same speed as the CPU's core, but that would be ridiculously expensive.
Since RAM tends to be accessed as chunks of instructions/data from adjacent addresses rather than individual bits scattered all over the place, designers add a relatively small amount of fast RAM as a "cache" to the CPU chip. Although the first time the CPU needs data, it has to wait for the slow main RAM, when that data finally arrives, a copy is saved in the cache. If the CPU next needs data from a nearby address, there is a high chance (~90%) that the data is in the cache, which means no delay in reading it into the CPU.
The bigger the cache, the more it can hold but the more expensive it is. Normally, current CPUs have a small primary "L1" cache which runs at full CPU core speed with low latency and a secondary "L2" cache with longer (slower) latency or slower speed, or both.
L1 cache sizes are typically on the order of 8-64K, while L2 cache sizes range from around 128K to 6MB or more. Today's mainstream CPUs have at least 2MB of L2 cache.
The bottom line is that L2 cache pretty much just acts as a buffer for the slow main RAM; as long as you have "enough", more doesn't make a big difference. A few MB is about "enough" these days.