May 24, 2016

Why do you need to know what PCI-e lanes are when building a computer?

Recently I started putting together a computer system that I want to use for at least a couple of years before it is considered “too” out of date, although nowadays a pc is out of date as soon as you power it on. Coming from using laptops most often I didn’t really have to care about the specifics of hardware communications or the frequency of ram modules, I just had to care about the surface specifications of the laptop to make a purchase.

Now don’t get me wrong the last time I built a pc from the ground up was around five years ago so I do know somewhat of what to look at component wise and that you cannot just mash components together like you want to. Recently something that really caught my attention was the concept of PCI Express lanes or (PCI-e Lanes). This comes in to effect between your CPU and you PCI-e components and how they will communicate to one another.

So firstly when you are buying the motherboard you have to look at the available lanes that is available to you. In my case I went with the “MSI X99A”, this board comes with a four “PCI-E x16” slots and two “PCI-E x1” slots.

(You don’t have to use all of the lanes, these are just the max potential)

What this should indicate to you is that in total you will need 66 lanes on your CPU to fully use the potential of all the interfaces at once. So second to the motherboard you have to decide on what CPU you want to use as part of your system. This will give you the max lane usage of your system when buying the other peripherals.

For example if you compare the “Intel Core i7 5820K with the “Intel Core i7 5830K”, they are very close in comparison yet the “5830K” is way more expensive than the “5820K” and this is because the 5820K only supports 28 lanes the 5830K support 40 lanes.

The more lanes your CPU has the more will be available when they need to communicate to the PCI-e components such as Graphic cards. The increase in lanes mean you will have higher transfer rates for the data that needs to go to the components, resulting in an increase of performance and keep the performance when adding multiple peripherals.