Desktop running dolphin slower than laptop

[Burrito]

Hello, a while back, I upgraded an old prebuilt PC, but it seems to be running worse than my old laptop, the only thing that my pc is worse in, is the fact that I am using ddr2 memory while my laptop is using ddr4, anohyter thing may be that I am using a modded xeon (E5460) which could also be a problem, this has been boggling my mind for a while, any help?

[Ivybridge11]

Hello, a while back, I upgraded an old prebuilt PC, but it seems to be running worse than my old laptop, the only thing that my pc is worse in, is the fact that I am using ddr2 memory while my laptop is using ddr4, anohyter thing may be that I am using a modded xeon (E5460) which could also be a problem, this has been boggling my mind for a while, any help?

That Xeon is over ten years old and gets absolutely creamed by any relatively modern CPU, including the one in your laptop, which if it uses DDR4, must be at least from the last few years.

[Burrito]

That Xeon is over ten years old and gets absolutely creamed by any relatively modern CPU, including the one in your laptop, which if it uses DDR4, must be at least from the last few years.

It also supports ddr3 1333 though, if I were to use that would it help?

[AnyOldName3]

I guess you've probably fallen foul of a couple of common misconceptions.

One is that more cores means more speed. This is only true in certain applications, as it's quite a lot of work to split some jobs between multiple cores, and sometimes doing so even makes things slower. As a human-scale example, think of writing a sentence - it's much faster for one person to write a sentence on a piece of paper than it is to somehow make two people do it. You could try swapping seats and the pen between each word, but the swapping would take longer than the writing. You could try each having your own pen and seat and just moving the paper, but it would still be slower. If it's a complicated sentence, then maybe you could plan out what it's going to say by discussing it, but the actual writing itself is fastest if one person writes all the words.

Another is that more clock speed means more speed. This is a little trickier to grasp, but the key thing is that a CPU doesn't necessarily do exactly one thing per clock cycle. Sometimes it does nothing, for example because it's waiting for some data to be loaded from memory, and that's when things like having more cache helps, as it means the CPU is more likely to be working on data it's got nearby. Other times, a CPU is doing more than one thing each clock cycle, such as when a vector instruction from an instruction set extension like AVX is used, and more than one piece of data can be worked on by one instruction. There are lots of other interesting techniques used to try and make CPUs do more things with the same number of cycles, but the end result is that a modern processor is much faster at the same frequency than an older one.

[Burrito]

I guess you've probably fallen foul of a couple of common misconceptions.

One is that more cores means more speed. This is only true in certain applications, as it's quite a lot of work to split some jobs between multiple cores, and sometimes doing so even makes things slower. As a human-scale example, think of writing a sentence - it's much faster for one person to write a sentence on a piece of paper than it is to somehow make two people do it. You could try swapping seats and the pen between each word, but the swapping would take longer than the writing. You could try each having your own pen and seat and just moving the paper, but it would still be slower. If it's a complicated sentence, then maybe you could plan out what it's going to say by discussing it, but the actual writing itself is fastest if one person writes all the words.

Another is that more clock speed means more speed. This is a little trickier to grasp, but the key thing is that a CPU doesn't necessarily do exactly one thing per clock cycle. Sometimes it does nothing, for example because it's waiting for some data to be loaded from memory, and that's when things like having more cache helps, as it means the CPU is more likely to be working on data it's got nearby. Other times, a CPU is doing more than one thing each clock cycle, such as when a vector instruction from an instruction set extension like AVX is used, and more than one piece of data can be worked on by one instruction. There are lots of other interesting techniques used to try and make CPUs do more things with the same number of cycles, but the end result is that a modern processor is much faster at the same frequency than an older one.

Well, damn. Is there nothing I can do?

[SCOTT0852]

Is there nothing I can do?

All you can do is upgrade the cpu in the desktop PC to something more recent. Don't forget to also give it a matching performance graphics card if you do.

[Burrito]

All you can do is upgrade the cpu in the desktop PC to something more recent. Don't forget to also give it a matching performance graphics card if you do.

Would overclocking help?

[JonnyH]

All you can do is upgrade the cpu in the desktop PC to something more recent. Don't forget to also give it a matching performance graphics card if you do.

Their profile says they're running a 1030 gddr5 - which should be powerful enough for just about everything, it being only a little slower than the 750ti that was generally recommended for 3xIR 1080p-level performance.