Thanks.
The general point is that (as I was told) slot a1 and b1 are slower Because... well, I'm not going to repeat it.
Once that 4 pieces consume more energy than 2 even with a larger capacity.
To quote : "Each memory chip consumes power, so having twice as many is twice the load to the motherboard's VRMs. DDR5 is unique in that the power supply for the memory is onboard the memory itself and no longer relying on power from the CPU.
Still, the memory controller on the CPU (or I/O die in the case of Ryzen) means that having four sticks means double the processing taking place in a small area and through the traces on the motherboard. Generally speaking this means that higher speeds are not possible. So if bandwidth is important you want the smallest number of larger sticks. If capacity is most important, you also want density. 196GB is possible using 48GB sticks, but you can expect this to be at 4800MT/s rather than a potential 6000-6400 for Ryzen
the bottom line is that four DIMMs presents twice the electrical stress on the memory controller, and thus, the CPU, as two DIMMs does. Twice the electrical stress also means, more heat, more electromigration potential, and a much higher chance of a reduction in the quality of signal integrity. Which is not to say that systems cannot utilize four DIMM configurations. They can, of course. But it is basically almost always a better choice, on consumer platforms at least, to use only two DIMMs, IF you can achieve the same desired overall capacity, because it's less stress on the IMC and it's generally also cheaper. A 32GB kit that consists of 2 x16GB is almost always going to be cheaper than a 32GB kit that consists of 4 x8GB, assuming that we are talking about a SINGLE matched set and not four individual DIMMs or two 2 x8GB kits. Doing THAT, opens you up to WAY more probability of compatibility issues among the DIMMs. That's another concern as well. The more DIMMs involved, the greater the chances that something is not going to want to "play nice together".
