Does Memory Performance Bottleneck Your Games?

[blazorthon]

[citation][nom]Angry Bellic[/nom]I don't see the meaning of people using 64+ GB of RAM, I mean I am a HARDCORE GAMER WHO ONLY PLAYS GTA IV AND GTA IV EFLC ( ) And I only got 16 GB of RAM whats the use of 64/32/128 GB of RAM, anyway?[/citation]

It's useful for some workloads, but yes, high RAM capacity for even above-average consumers such as mid-ranged and high end gamers is generally a waste of money. By high, I'm referring to beyond 8GB. 16GB can be reasoned for if you do more than gaming alone, but anything more than that is generally only useful in some servers and workstations.

 

[ericjohn004]

2133 RAM is of great benefit. Maybe not in games. But when I went from 1600 to 2133 I saw a 700 point performance gain in 3DMark11 and a 5% overall PC speed gain in Cinebench and many other programs. It gives you about as much CPU performance as a .15Ghz overclock. And if you already maxed out like I am that sure is welcome.

 

[Achoo22]

[citation][nom]Crashman[/nom]...then you won't be buying all this other expensive hardware...and anyone who has that much money to spend on gaming can probably justify a few dollars extra for RAM.[/citation]

This is exactly the kind of backwards thinking that causes folks to defend expensive cases that offer no meaningful improvements over cheaper ones. "Nobody is going to spend X dollars on gear and put it in a case costing Y." If you're using a person's wealth to justify premium pricing, something is wrong.

 

[Crashman]

[citation][nom]Achoo22[/nom]This is exactly the kind of backwards thinking that causes folks to defend expensive cases that offer no meaningful improvements over cheaper ones. "Nobody is going to spend X dollars on gear and put it in a case costing Y." If you're using a person's wealth to justify premium pricing, something is wrong.[/citation]Yours is exactly the kind of backwards thinking that people use to justify doing nothing at all. The improvements are fairly large in some applications and if you give me the "difference between 100 and 120 FPS" argument I can just as easily toss back the "3D" argument.

 

[phraun]

[citation][nom]sempifi99[/nom]No, you are a troll for assuming I was lying. And saying 2004 is not far off, the computer was built in 2005 and it has an updated graphics card... let me list the specs for you below. Motherboard: Asus K8N-DLCPU: Dual Opteron 285Ram: Quad Quimonda 2GB PC3200R Gpu: EVGA GTX 680[/citation]

That system disturbs me on so many levels... It's like you added NOS and twin turbos to a 1980 Nissan pickup, but without the comedic value.

 

[coolronz]

well from my own experience, on a i7 3820 and ASUS P9X79 Pro, going from Corsair Vengeance DDR3-1600 cas10 to TEAM-X DDR3-2133 cas10 my rig was noticeably more responsive and benched about 10% faster.. not to mention ASUS auto-overclock finds 4.7GHZ @1.38v.... altho I have a 2x crossfire HD7970 setup, its maxed out by my 5760x1080 eyefinity display.. gamewise for me didn't make a difference, but for general multitasking its much better, from my experience....

 

[Hindesite]

So what I'm reading here is that my old DDR2 800 might be holding me back some?... lol

 

[blazorthon]

[citation][nom]Hindesite[/nom]So what I'm reading here is that my old DDR2 800 might be holding me back some?... lol[/citation]

In some ways, possibly.

 

[NightLight]

my opinion: new rig, fastest available ram...
that said, my rigs last about 5 years now, they used to just last 1.

 

[redice611]

Did anyone notice the funky numbers in Battlefield 3 where the 4800x900 yielded better results than the 1920x1080?

 

[knowom]

Memory bandwidth and latency has a bigger impact for audio and video work it can make a rather big difference for maintaining low midi latency without drop outs occurring where you need all data bits to be processed and arrive on time in the right order.
Memory latency and bandwidth generally has always only offered up minor improvements for gaming itself density on the other hand could much more regularly offer better improvements far worse being memory starved then having a bit lower bandwidth/latency for gaming in my experience.
A nicer article would have been to try and benchmark improvements when using fancycache or supercache for a dedicated drive or partition with a game on and see how if at all memory latency and bandwidth impacts that as both latency and bandwidth should provide some improvements for those memory cache based software bandwidth in particular is highly favorable for them.

 

[Adroid]

I would have really liked to see a Cas 11 result to see how much more noticeable the game performance would be compared to the Cas 9 at a given frequency. Also, 16gb vs 8 gb would be a nice touch.

 

[no1spank]

How effective the ram is depends a lot on the quality of the memory controller in your particular chip. If you are not overclocking then get the speed matched to the chips rated mem controller with the lowest latency.I overclock a lot of AMD chips and found that a stick won't overclock with one processor but will fly with another (all other parts are the same).Higher rated memory does solve this problem and make higher stable overclocks a lot easier on a wider variety of processors.As a rule it is best to get the fastest memory if price is similar, bear in mind that dropping the speed of faster memory doesn't automatically mean it will run tighter timings and slower memory won't always run faster if you loosen the timings. As I say your particular processor has a bigger say in it than you may think.

 

[iam2thecrowe]

bit of an old thread to respond to, but i would say, if your running a game that is maxing out the vram of your video card, it spills into system ram. so if you have faster ram in this scenario you would surely see a benefit from it. Also BF4 gains quite a bit of fps from faster ram going from benches floating around the web, again could possibly be that whatever vid card they used to test this is just running out of vram.

 

[AnEwG]

This testing rig is monstrous.