Overclock.net banner

[Benchmark] ECC Registered vs Regular Ram [Rainbow Six Siege]

14K views 24 replies 8 participants last post by  Synoxia 
#1 · (Edited)
Xeon E5-2689 @ 3.3 ghz

64 GB ECC 1333mhz DDR3 RAM
vs
32 GB Regular Gaming 1600mhz DDR3 RAM

Fun fact: ECC Registered RAM at same mhz is actually faster than regular non-ecc ram, and faster than ECC Un-Buffered RAM modules.

In this test you can see about 2-6 fps loss to gaming ram, because ECC is at 1333mhz, and gaming ram is at 1600 mhz.

As long as timings are good on ECC RAM and it's in same mhz frequency as regular ram, ECC will win gaming tests.

BUT. ECC RAM runs hot, and that could be a deciding factor for most. ECC RAM is much cheaper.

ECC RAM is usually the fastest for gaming if comparing 1600 mhz to another non-ecc/ecc unbuffered counterpart. If you overclock regular ram above 1600 mhz (with same timings as ECC RAM), then you will have a winner using regular RAM [for gaming purposes]

ECC provides error correction for data sensitive material, and overall is much stable than regular ram. Where your computer will bluescreen using regular ram, ECC ram is likely to correct the error. That is of course if it's not CPU/hardware related issue.

Another great info on this:
https://www.techspot.com/article/845-ddr3-ram-vs-ecc-memory/



Here is another test same frequency (1333mhz), ecc vs non-ecc

I tested old Lenovo S30, now the other non-ecc version is on X79 MSI motherboard. So I was able to downclock non-ecc sticks I had
 
See less See more
#2 ·
Do you need a gaming chair in order to use gaming RAM or can you have one without the other?
 
#3 · (Edited)
Sometimes I wonder if you just troll these forums, just for trolls.
You do know there is such thing as gaming ram per definition of what they say, that's what they market it as, because it runs in performance mode, at better timings and faster mhz.

I actually opened this thread happy, thinking someone contributed to it, but instead I see this.

Hard to say who you are ridiculing, me, or the concept. Either way, it's bad taste, bro.

While I am just trying to educate people. Especially the ones who google answer all day long, actually learning.
 
#4 ·
Although I appreciate the post, could you also do a comparison between the 2 sets, where :

They're running at same frequency and
They're running the same capacity

This could give what the real world difference is between the 2 sets.

Also don't mind Pook. It's just the bandom leaking.
 
#6 ·
I keep on coming back to OCN because of users like The Pook and a few others.

Slightly higher bus transfer for Slightly higher latency. This isn't DDR2 FB-DIMMs

I have a X79 rig with SAMSUNG PC3-12800R DDR3-1600 RDIMM and 990FX with Micron DDR3-1866MHz CL10 UDIMM


ECC ribbed for her data sensitive material. You can overclock ECC memory.
 
  • Rep+
Reactions: The Pook
#7 ·
You have normal motherboard, I have Lenovo S30, which has no way to change clocks, multipliers, or base clock.
To what you said about Pook...well, I don't. I respect people who have respectful input, not disrespect. Considering how I carry myself in general, and on these forums as well. I try to help people, not ridicule them. If I had to deal with this attitude on the daily, I wouldn't be here.

Seniority doesn't condone that type of behavior. To me what matters is how you treat others, only after what you know. Besides, I was expecting some input from people. Not sly backhanded comments.

Like we know each other.
 
#8 ·
Guess I forgot the internet was for serious business only.
 
#9 ·
There is a thing called tact, Pook
When you spend time taking a benchmark, then cropping stuff properly, syncing first frame of each video to make it easy to see for people, then render, then upload, then create text gathering of the data gathered, to be greeted with "Cool gaming ram, bro"...

In other words, don't be an a-hole

If you don't value my work, at least be respectful. I think in my experience with you, I showed nothing but respect.

There is a place to be funny, and then there is a cheap thrill of getting a laugh for no reason, then actually taking time to type out what you thought, while ridiculing the person.

Any more analytics I must type, to clarify your attitude.

I like this forum in particular, because there is knowledge here. People with vast knowledge that have helped me build my computer, helped me pick out right stuff, and generally made my life easier. I saw so many guides here, that I read and appreciated. You know that phrase: "If you don't have anything good to say, don't say anything at all"

The problem with a rude person in general when he is rude, he doesn't see anything wrong with it. To you it's a joke, to me a discount of my work, because you found a "funny"

Internet is no different than real life. Only in real life I would just shake my head at it, because you would see my response, with your eyes.
 
#10 ·
So I'm not allowed to "do a funny" because you spent time on something?

:laughings

thanks for the life lesson.
 
#12 ·
This is interesting though that's quite old hardware by now; it sounds like this is on an OEM computer and all you can change is your DIMMs?

I wish ECC RAM were more widely usable, just to play with.
 
#13 · (Edited)
Your CPU is chogging the GPU, RAM frequency won't matter at this time. Score difference is more like a margin of error as differences are not consistent and ECC RAM happens to have higher scores.
 
#14 · (Edited)
Yes, it's OEM Lenovo S30 in first test.

Why not consistent, the difference is clearly there, and it's repeatable. And yes, ECC ram is better for gaming than regular ram, as I said in first post. It is slightly bottlenecked, slightly, to the point I get almost zero stutters in games, and that's on 2012 chip vs 2018.

Here is another test same frequency, ecc vs non-ecc

I tested old Lenovo S30, now the other non-ecc version is on X79 MSI motherboard. So I was able to downclock non-ecc sticks I had


What I know is that different motherboards have different response in ms. I know because I would test same CPU on different motherboards, and HP z420 had as much as 3 ms difference in Intel Burn in tests, while Lenovo S30 pulled ahead by 3 ms.

I did notice the same trend when I was testing i9-9900k across different motherboards, about 10 to be exact. Every board showed less or more ms delay in rendering frames and Intel Burn in test was different, with the best board hitting about 6.6 seconds in Intel Burn in.
 
#15 ·
If you say so. Can you check timings in AIDA64 (motherboard - > Chipset) and also run a baseline memory test for both kits there, please?
 
#17 ·
There's more to timings than primaries.
 
#19 ·
Good luck finding tertiaries and secondaries in datasheet.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top