There are already a few questions about people looking to upgrade their RAM to something faster than the CPU or motherboard supports (e.g., this or this), and it seems that the RAM will always be downclocked to the lowest maximum speed of the three components. This makes sense.

Why, then, do some computer manufacturers use faster RAM than supported by the CPU? I found this out when helping a friend check out laptops, and then looking at my own laptop.

  • Friend’s Laptop: (New) Dell Inspiron 15 5000: CPU is i7 8565u, which supports up to DDR4-2400, but laptop includes DDR4-2666.
  • My Laptop is Similar: It has a i7 7500u, which supports up to DDR4-2133, but includes 12GB of DDR4-2400. I can see that it is clocked down to 2133MHz using dmidecode.

Maybe more interestingly, on my laptop 4GB is integrated, but the other 8GB is in a removable SODIMM. If this motherboard were mass-produced, I could see how they incorporated a higher-clock RAM if they wanted to support more systems without a design change. But surely the OEM laptops could have had customized the SODIMMs to match the processor and reduce costs?

I know clock speeds aren’t the only factor important to RAM speed (there’s also CAS latency), but I don't imagine companies to choose higher-clock frequencies because of these other properties.

In other words: Why do manufacturers produce laptops with RAM at clock speeds their CPUs can’t use, when it seems more economical to use lower-clock, (presumably) cheaper RAM?

  • 5
    I can't tell you WHY companies this specifically, but it is likely economics... It would likely be cheaper and more efficient to purchase, inventory and allocate 10,000 pieces of DDR4-2400, than 5,000 DDR4-2400 and 5,000 pieces of DDR4-2133. For example, look at Newegg and in purchasing 2 units of each, the cost difference is $1 or less per pair at retail, in some cases they are the same, and 2400 is even lower in a few cases... Why bother two identically priced items when one will do the job in both cases. Less organization, chance of mistake, etc. the end result is one type is cheaper than 2 – acejavelin Jun 29 '19 at 2:18

The rated clock speed of RAM is not the clock speed it operates (or must operate) in, but its maximum supported clock speed. Thus there is no "clocking down" per se, it's just run at a lower clock speed than it could.

It likely doesn't make sense to produce RAM that would not run at the currently most popular speed, as the production volume itself will push the costs down more than decreasing the quality for a small part of the batch would. This translates to lower prices as well, especially for customers who buy in bulk -- they likely get the higher-rated RAM cheaper than the lower-rated.

| improve this answer | |
  • So it is just an money issue! Thanks to you and @acejavelin for your inputs. (I'm not good at thinking about the greater business scheme, it seemed that the cheapest parts would mean lowered costs.) – Jonathan Lam Jun 29 '19 at 4:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.