Effects of mismatched ddr4 sodimm

I just got the xps 15 9550

It comes with 2 4 gb ddr4 sodimms

It costs about the same to get 2 8gb sodims or one 16gb sodims

So suppose I only have 100. Isn't getting a 16 gb sodim and having 20 gb now and 32gb later better than 2 matching sodims

Is there a speed difference and what would it be?
Thanks for your answer if you know.

(or said a different way if you don't know please don't answer)

 

Depends on the notebook/platform/BIOS and possibly cpu.

Depends on the workflows you depend on.

Depends on the Windows O/S version you're using.

Depends on the software used and the options within that software.

Will it work? More than likely.

Speed difference? Not more than 5%, overall.

I would say; suppose you save for a few more weeks and install 32GB at once, instead.

 

Ok that is definitely a I don't know answer please don't answer again

This is a i5 skylake ddr4 It's Windows 10.

If there is an actual owner with information to go on or a link to a ddr4 dual channel ram test please feel free to answer

No educated guess thanks .clear.

 

Actually, he did provide an adequate reply.
How RAM mismatching impacts performance varies from one software (and system) to the next...
Where you are most likely to see the impact of dual-channel (synchronous) RAM is likely in integrated graphics performance and software that relies on RAM speed and bandwidth.

If you were using an AMD system that had dual-channel enabled, IGP performance (and likely CPU performance) would jump noticeably - but this is mostly confined to AMD APU systems that rely a lot on RAM bandwidth and definitely operate better in dual/quad channel configurations. But since you don't have an AMD APU, the point is moot (I was only illustrating a comparison).

Intel IGP's might benefit from dual-channel RAM, but since your laptop has a dedicated GPU (960m), I don't see how mismatched RAM would affect performance in your case.

If you are thinking on getting 32GB RAM eventually, you might be better off getting it immediately if you can afford it, or as you said, simply get yourself 16 GB now, and then later upgrade with another matching 16GB stick that has same specs.

20 GB RAM should be enough for the time being until you can upgrade to 32 (more RAM certainly doesn't hurt).
I'm still using 8GB on a laptop in my signature and I don't see issues - though this also depends on which software you use, and how often.
Using 3d Studio Max in addition to modern software for instance (such as Google Chrome) can easily push RAM use pretty high up... in which case, 16GB and even 32GB would definitely be of benefit.

But as I said, I don't think your system would experience any noticeable performance loss with mismatched RAM.

 

Okay, get synthetic tests if you want.

Don't provide any additional information (as asked).

And insult the person trying to help.

You're asking for a speed difference and I told you (~5%), but that will also depend on your software you run too.

Btw, if the ram tests you're asking for were available, you'd have found them by now.

 

I've provided all the information necessary please don't answer if you don't know from your personal computer or a direct link

 

I hear you as far as tests.
I have a ddr3 test that shows there is no difference of even 1 percent.

 

If there is no like pc magazine type test I'm sure a laptop owner has tested but if not I'll go on my educated guess

 

@stamar just get that 16GB stick already. It will work. You will have 8GB running dual-channel and 12GB single channel. Theoretically, the difference in some workloads may be up to 10% once you're using over 8GB - but in reality it's couple % and you won't notice it.

 

Ya I can't think of a reason to dual channel 2 8gb for 80 dollars when I can go to 20 gb for 100

And then I'm ready for 32gb just have to get the same brand

But because ddr4 Is new someone might have some news about it. I haven't seen any tests on it

 

You haven't provided anything resembling a workflow and/or indicated the software used.

See:
http://www.legitreviews.com/ddr4-me...nding-the-best-ddr4-memory-kit-speed_170340/6

See:
http://forum.notebookreview.com/thr...p-screenshots-appreciate.699186/#post-8976613


The first link shows the best RAM for the Skylake platform. There are more articles like that if you search long enough.

The second link shows that even in platforms that are not capable of manually setting memory timings in the BIOS, an 'optimized' module will still perform faster than a default one (as per spec's).

I still suggest you buy the biggest, fastest pair of DDR4 RAM modules you can save for and/or afford.

Sticking with stock DDR4 2133MHz RAM (at 16, 32 or 64GB+ capacities) is not what I would be spending my $$$ on.

At least not if it meant just saving for a few weeks longer. And, the extra RAM capacity was needed in my workflows, 'now'.

Good luck.

 

Method to test performance gain of single channel vs dual channel removed at request of OP.

 

Quoting because I asked.

Running a 'bm' won't tell you what is faster for actual workloads (except for what scenario is faster to run that bm).

 

Because he asked....

Ibdont have a laptop to test I'm shopping so I am asking for someone else to do it. You seem totally and completrly detached from the discussion your comment has no point ....?

I'm asking you to run a test rocket scientist. No kidding if I run a test I can figure it out..... After I already buy the ram.
That way I can save time and not post on a forum.......?

 

Of course, it demonstrates a prime95 type workload if however the OP runs say "typing with Mavis" then he/she shouldn't expect it to make any difference to his/her typing speed.

 

Noted, from now on will not waste your time or mine with replies to your questions.

 

Thanks

Not that I want to argue with people but everyone here wants to be a know it all but the threads are always filled with weird bs.
So it's frustrating for me because I'll find and state the correct answer myself in the thread for the next person who has the same question but then they read posts on weird stuff that makes no sense.

 

Which reminds me that's all desktop ram in the article
But it has good info just to see which ones are on top of their ram game

 

It's not the RAM they're testing; it's the platform (Skylake).

See:
http://www.hardocp.com/article/2015...76700k_ipc_overclocking_review/6#.VsF81Wb2Z9A

See:
http://www.anandtech.com/show/9483/intel-skylake-review-6700k-6600k-ddr4-ddr3-ipc-6th-generation/7

See:
http://www.corsair.com/en-ca/blog/2015/september/ddr3_vs_ddr4_generational

See:
http://www.tomshardware.com/reviews/skylake-intel-core-i7-6700k-core-i5-6600k,4252-3.html


From the last link above:

What does this show us?

Exactly what I told you before;

"I still suggest you buy the biggest, fastest pair of DDR4 RAM modules you can save for and/or afford."