Religeous attitudes towards memory kits

Valentin Tihomirov

Honorable
Apr 10, 2013
71
0
10,630
Yesterday, I asked why are the memory kits are 30% more expensive than discrete parts. The answer was that kit suppliers provide replacement in case of incompatible memories. But, in addition, the proponents of this idea also say that the compatibility is ensured by testing. This is what I do not understand. I do not understand why do not they listen be when I say that memories are not attached to each other -- they are attached to the mem controller instead. There are physical lines which connect a memory slot with a respective channel of the memory controller. This is peer-to-peer connection. The other sticks are attached to another channel. This means that memories are connected only to the controller. Furthermore, there are different (e.g. ganged and unganged multichannels), which impose different requiremens on the memory kits. Why do you need to test the memories against each other rather than against the controller that they are attached to? There is not answer from the religous people. Religous people can only shine erudition informing me that mem controller is inside my CPU (like it is not on the my board) and keep bradcasting that this allows the memory kit supplier to test the memories. How?? They broadcast that memories are complex, have billions of tiny transistors, and, therefore it is difficult to find compatible modules. Well, ok, but shouldn't the modules be compatible with the controller, CPU, which they are attached to, with the board layout that varies the delays immensly? No answer. The religous people keep broadcasting: "the memory sticks must be tested against each other". "the memory sticks must be tested against each other". "We have not time. We are not payed for helping you. Be happy for what we say to you. ". "the memory sticks must be tested against each other".


An these broadcasts that 1333 MHz memory works better with 1333 memory drive me nuts. What is the ground of your claim? The universal approval from the community? I know that you will recieve it making such broadcasts. But, do you know that when you plug multiple sticks into one channel, the capacitive load is increased? Do you know what is the capacitive load or your judgement is based on ungrounded gossips and numerology (1333 better matches 1333 than 1666). This is intuitive, isn't it. The sticks have the same characteristics. They have a higher chance of being compatible, isn't it? However, if you are familiar with electronics, it is intuitive that 1666 memory is faster, and, thus, has less capacitive load, and thus has lower performance degradation effect, no matter which module it is matched with. That is, if you are familiar with electronics, 1666 better matches 1333.

I resorted to tomshardware because I felt like this is a place where people test things and analyze the performance bottlenecks. However, the advices in the forum seem to be a flood of ungrounded misconceptions.

I do not accept the answers from broadcasters. I want to know why 1333+1333 in the same slot is better than 1333+1666 and how can you test memory compatibility if you do not have my motherboard from the technical point of view. Be critical to what you say.
 

Valentin Tihomirov

Honorable
Apr 10, 2013
71
0
10,630
Tradesman1, this is what I am talking about. Obviously you have no idea how systems work and all you have to tell me about the electronics is your religious argument. You have no electronics argument besides religious manipulation and my demand of technical reason makes you angry. Thank you for confirming your expertise. Keep manipulating.