Overclock.net banner

Battlefield 4 RAM / Memory Benchmark

39K views 78 replies 34 participants last post by  TiezZ BE 
#1 ·
Dear community,

after I read this article ( http://www.corsair.com/en/blog/bf4-loves-high-speed-memory) , I was sure, because Battlefield 4 will be my main game for the next years: I need to test this out.

Marketing-Gag or truth?

Components

Intel Core i5-2500K S1155 Box
Powercolor HD 7950 V2 mit AMD_Catalyst_13.11_BetaV1
ASUS P8P67 LE (Rev 3.0)
2x1TB Samsung SpinPoint F3 HD103Sj 7200 U/min
3TB FANTEC DB-ALU3e 7200 U/min
128GB Crucial M4 MLC SSD
LG BH10LS30
Corsair TX650 V2 80+
Sharkoon Tauron Green Edition
Mugen 2 Cooler
Windows 8.1 Pro 64bit
(BenQ Xl 2411T 144hz)

overclocks:

i5 2500k @ 4,8ghz
Powercolor HD 7950 V2 @1150 Coreclock with 1,17v , @1400 Memoryclock with 1.58v

RAM:

Kingston ValueRAM DIMM 2x4GB, DDR3-1333, CL9-9-9-24 1.5v(KVR1333D3N9/4G)
G.Skill RipJawsX DIMM Kit 2x4GB, DDR3-2133, CL9-11-10-28-2N 1.65v(F3-17000CL9D-8GBXM)

Testscenario:

Domination Siege of Shanghai, on the roof (see video)

Private Server 1/32 Players, "Scooby Snacks BF4 PCW/CW" http://battlelog.battlefield.com/bf4/en/servers/show/pc/d9788992-49d3-403a-9c9d-b3f1dbce62e1/Scooby-Snacks-BF4-PCW-CW/

Video of the test scenario (short cut of the benchmark, benchmark is 180s long, video is lagging because my hard drive is not fast enough for fraps):

http://youtu.be/GGUlkTA-JII

For every RAM 2 runs.

Ingame settings:
xfzjw47ajsyj.jpg

Results:

1333mhz:
ywywixss62x2.png

2133mhz:
53v42khebur.png

1333mhz vs 2133mhz:
4otlxh5y6iu.png

Conclusion
  • less framedrops or less incursions
  • FPS increase 58%, RAM mhz increase 60%
  • -> in cpu-limited situation the fps increase is nearly the same as the ram mhz increase
Extra test on Ultra:
(only one run)
G.Skill 2133mhz vs 1333mhz

Settings:
v2iuhhsoxf7.jpg

Results:
xctovy4o9ljq.png

  • 13% fps increase overall, 9 more minimum fps, 28 more maximum fps
  • ->even in GPU-limtied situations there is still a good overall increase
2.Extra test on Ultra:
(only one run)
G.Skill RipJawsX DIMM Kit 2x4GB, DDR3-2133, CL9-11-10-28-2N 1.65v
vs.
G.Skill RipJawsX DIMM Kit 2x4GB, DDR3-2133, CL9-10-10-27-1N 1.65v
482z449lqsau.png

Downloads:

Excel sheet for low settings:

Vergleich 1333mhz vs 2133mhz.xlsx 23k .xlsx file

Excel sheet for ultra settings.:

Ultra Vergleich GSkill 2133mhz vs GSkill 1333mhz.xlsx 19k .xlsx file

Excel sheet for 2133mhz vs 2133mhz OC (Ultra Settings):

Ultra Gskill 2133mhz vs 2133mhz OC.xlsx 19k .xlsx file


I hope this was helpful for some of you guys. For me the upgrade was totally worth it.

Sincerely,

GamingWiidesire
 

Attachments

See less See more
10
#3 ·
Thanks for doing this, much more reliable than Corsair testing their own RAM to show how much better this and that is etc.

I got almost the same components as you in the tests so I might just go ahead and upgrade to some cheap 2133MHz-RAM, mainly because of these tests
smile.gif


+REP.
 
#4 ·
If it is not too much of an issue, will it be possible to test 1600 MHz and 2133 MHz? You could probably just install the 2133 MHz kit and down-clock it to 1600 MHz.
 
#7 ·
I don't know why I am getting the same results for 1600MHz vs 2800MHz

1426418_659180650781018_1534015831_n.jpg


2560 X 1600 with single GTX780 Ultra settings
 
  • Rep+
Reactions: ricklen
#9 ·
Quote:
Originally Posted by massman View Post

DDR3-1333 is rediculously slow anyway, no wonder you're getting a massive boost.

The highest JEDEC standard is DDR3-2133. You should start comparing from that baseline, and then work your way up to DDR3-2666 or so.
Well, 1333 and 1600 is quite widely used, so it's of interest... especially when past general sentiment was that RAM speeds didn't massively affect frame rates.
 
#10 ·
Quote:
Originally Posted by 98uk View Post

Well, 1333 and 1600 is quite widely used, so it's of interest... especially when past general sentiment was that RAM speeds didn't massively affect frame rates.
exactly. I also bought 1333mhz ram, because all told me "it doesn't affect framerate, or just by 1-2%"
 
#12 ·
Quote:
Originally Posted by anubis1127 View Post

When were you told that? 1866Mhz has been the "sweet spot" for gaming for the last two years or so.
I bought my 1333mhz ram 2,1 years ago along with my i5 2500k. Back then everyone on "computerbase.de" said go for the cheapest ram, it doesn't matter
biggrin.gif
 
#13 ·
Quote:
Originally Posted by 98uk View Post

Well, 1333 and 1600 is quite widely used, so it's of interest... especially when past general sentiment was that RAM speeds didn't massively affect frame rates.
You make a valid point. I'd say there's an argument to be made not to go for DDR3-1333 altogether. The absolute price of a DDR3-1866 kit is not that much higher anyway.

On topic, AnandTech did a similar test a while back: http://www.anandtech.com/show/7364/memory-scaling-on-haswell/7. Not that much difference there. Might be just this game.
 
#14 ·
Quote:
Originally Posted by owikh84 View Post

I don't know why I am getting the same results for 1600MHz vs 2800MHz

2560 X 1600 with single GTX780 Ultra settings
GPU bottleneck?

Try 1080p
 
#17 ·
Quote:
Originally Posted by JSTe View Post

GPU bottleneck?

Try 1080p
but the Corsair blogger used 1920x1200 and 5760x1200 and the differences are so significant.
btw he is using GTX780 in SLI mode while I am on a single 780
 
#18 ·
Quote:
Originally Posted by owikh84 View Post

I don't know why I am getting the same results for 1600MHz vs 2800MHz

1426418_659180650781018_1534015831_n.jpg


2560 X 1600 with single GTX780 Ultra settings
I'd like to see a bit more testing from you with 1600 vs 2800. You should see a much bigger difference. Can you retest and maybe check to see where your bottleneck is at the time of testing?
 
#20 ·
Quote:
Originally Posted by Gallien View Post

I'd like to see a bit more testing from you with 1600 vs 2800. You should see a much bigger difference. Can you retest and maybe check to see where your bottleneck is at the time of testing?
alright I will use another scene and do a longer test.
 
#21 ·
Quote:
Originally Posted by owikh84 View Post

alright I will use another scene and do a longer test.
OP is using an empty MP server by the way... one assumes the memory/CPU calculations are different to those done in SP (in case you were benching in SP)... things like networking, anti cheat etc...
 
  • Rep+
Reactions: owikh84
#22 ·
Quote:
Originally Posted by TiezZ BE View Post

I've got a question because I can't find them in the QVL list.

I just saw these (used):


(Info on the memory module is in the pic)

Will they work on my:

Asus P8P67 PRO?
My RAM wasn't in the QVL list on the Asus website, it wasn't even updated for 2 years. However on the G.Skill website I found a compatible Mainboard list including my mainboard. I'm on my mobile so I can't give you the link right now, but you could search on the website on your own.

Sent from my GT-I9505 using Tapatalk 4

edit:

yeah the ram is compatible:

http://www.gskill.com/en/product/f3-17000cl9q-16gbzh

P67
  • Maximus IV Extreme
  • P8P67
  • P8P67 (REV 3.1)
  • P8P67 DELUXE
  • P8P67 EVO
  • P8P67 LE
  • P8P67 PRO
  • P8P67 PRO (REV 3.1)
  • P8P67 WS Revolution
  • P8P67-M
  • P8P67-M PRO
  • SABERTOOTH P67
 
#24 ·
Quote:
Originally Posted by rivaxstorm View Post

My RAM wasn't in the QVL list on the Asus website, it wasn't even updated for 2 years. However on the G.Skill website I found a compatible Mainboard list including my mainboard. I'm on my mobile so I can't give you the link right now, but you could search on the website on your own.

Sent from my GT-I9505 using Tapatalk 4

edit:

yeah the ram is compatible:

http://www.gskill.com/en/product/f3-17000cl9q-16gbzh

P67
  • Maximus IV Extreme
  • P8P67
  • P8P67 (REV 3.1)
  • P8P67 DELUXE
  • P8P67 EVO
  • P8P67 LE
  • P8P67 PRO
  • P8P67 PRO (REV 3.1)
  • P8P67 WS Revolution
  • P8P67-M
  • P8P67-M PRO
  • SABERTOOTH P67
Thx!

Was also browsing for more info.

Is it a problem to use these 1.65V sticks on a sandy bridge CPU (2500k) for a long period??
"DIMM voltage below 1.65V is recommended to protect the CPU" says the manual of the mobo.
But saw here on OCN and other websites some statements that up to 1.65V RAM shouldn't be a problem on a sandy bridge system.
 
#25 ·
Well I'm running my i5 2500k at 1,47v on load since 2 years, that's also not recommended. In my opinion and what I've read 1,65v for the ram on sandy bridge should be the maximum, you shouldn't go higher
 
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