There is literally no difference in gaming or everyday use between DDR3 and DDR4.
None at all.
Please shill and justify the extra money you spent for zero performance increase below.
>>55298672
Nice meme.
http://www.techspot.com/article/1171-ddr4-4000-mhz-performance/page3.html
https://www.youtube.com/watch?v=Er_Fuz54U0Y
>>55298672
There is literally no price difference any more either
GET THE FUCK OUT TO /V/ YOU FUCKING UNDERAGE COCK SUCKER
>>55298672
Fuck off >>>/v/
>>55298697
But there's only DDR4 on that graph anon, how can you illustrate that it's better than DDR3 when it literally does NOT compare the two?
>>55298754
So DDR4 speeds matter but DDR3 speeds somehow wont...?
Are you retarded?
>>55298754
Because you'd need to be completely tech-illiterate to not understand how memory bandwidth and latency works, anon, and think that the number after DDR makes a difference to it.
You're not tech-illiterate... are you...?
Nice made up graphs, lol.
ddr4 automatically overclocks
>>55298697
This is a graph AMD owners use trying to prove they can squeeze a decent amount of fps out of an inferior gpu. Nice try.
No difference in DDR3 and DDR4 performance in any activity save serious mining and crunching operations. Still waiting for any legitimate counter argument.
I'll probably be waiting for a long long time, lmfao.
>>55298697
Fall out 4 is a giant unoptimized turd. RAM frequency should NOT effect FPS. This is a sign of system RAM being used for things instead of faster vRAM on the dedicated graphics card.
>>55298956
Might as well be a buzzword, since """auto overclocked""" ram doesn't provide any performance boost in any game.
hijacking this thread for a question:
does dualchanneling actually do something?
would I notice a difference between 2x 4GB in dual channel vs. 8gb single bar?
>>55298972
That graph has nothing to do with AMD. Bethesda simply can't into optimizations and seem to be using system RAM as vRAM for some retarded reason.
>>55299011
This.
>>55299004
For 90% of things it does jack shit. 9% is for AMD users with an APU and no dedicated graphics card and the 1% is for benchmarking.
In fact having a single stick of 1333 MHz DDR3 is good enough for someone not using an APU or who cares about benchmarks.
>>55298776
Again, there is zero real world difference in gaming or any other non specialized task.
Doesn't matter how many graphs you post. There is no difference besides $$$ and epeen size.
ddr4 is cheaper
>>55299072
lol
What difference does it make? Why are you so concerned with what other people buy. I have a DDR3 desktop now but the next one will certainly be DDR4, if for nothing else than the ability to use more memory. I have a quad core +HT ThinkPad with 32gigs of memory. Right now I'm running 3 VMs on it to prepare for an upcoming pentest. I've given all 4 OSs 8gigs of memory and one or two cores. It would be great if I could have more memory and run a few more VMs. I'd need ~two more to saturate my gig NIC doing what I'm doing. I already Max the cores out for a few mins at a time but unused CPU headroom is wasted potential when you are time sensitive.
I do even more with my desktop but at home I've got an ESXI cluster with half a TB in it, in my garage.
>>55299112
Not him, but its a $1 difference on that particular brand/model
>>55299112
o-o-okay? Freeshipping means its cheaper to get ddr4...
>>55298672
"I'm to embarassed to admit I don't know the difference, please discuss so I can listen in the corner while I count my pennies"
>>55299112
>>55299165
>gaymur RAM
Into the trash it all goes.
Faster RAM, be it DDR3 or DDR4 makes a difference in CPU bound fps.
Just by overclocking my 1600MHz set to 2400MHz makes a massive difference in GTA 5.
>>55298697
wew m8 you can´t read for two shits, it´s clearly stated that the difference you see here is made with crossfire/sli cards because at 2+cards ramspeed bottlenecks they SPECIFICLY say that @ one card it doesn´t matter if you use 2400mhz or 4000mhz
~10%-20% difference on 3/4 games tested in this scenario between 1600 Mhz DDR3 and 2400 Mhz DDR4
>>55298672
Define every day use.
RAM speeds can have significant impact on graphics performance when using an integrated GPU.
>>55299356
Between maximum FPS. Seeing how little the average changes, that's practically worthless.
>>55298672
Nice meme. Over 3000mhz it does actually make a difference, but only like 2-3fps. I'd you want a new system, you are going to get ddr4 no matter what. Might as well get the best performing product at the time. The higher binned ram is not much more anyway.
>>55299411
>Between maximum FPS. Seeing how little the average changes, that's practically worthless.
The charts are average fps.
Windows 10 boots 350ms faster with DDR4 compared to DDR3
>>55298697
>Fallout 4 as a benchmark
>No listed DDR3 comparisons
Did you even TAKE math classes?
>>55298697
All I see is that Fallout 4 is terrible. Ram speeds should not effect FPS this drastically if programmed correctly.
>>55302883
Can't argue with that, apparently Bethesda doesn't know how to use L3 cache effectively.
But that doesn't negate the massive improvement that DDR4 is over DDR3.
>>55302976
Sure, DDR4 is objectively superior, but if there is no practical gains then there is no point yet.
Intel told me to.
>>55301707
I wasn't aware that this was an /g/ humor thread.
>3600mhz DDR4
>15-15-15-35 timings
>8.3ns~ true latency
>at 57.6GB/s
DDR3 never got close to that. Its true that not many every day programs would benefit from this, but your every day programs don't benefit from a Skylake i7 over a Sandy Bridge i3 either. If your benchmark for system performance is a web browser then you're never going to care about hardware improvements.
>>55303857
What is the density on those?
I have 3200MHz 16-18-18-38. 4x8GB
I get the following in AIDA64 memory test
>Quad DDR4-3200 16-18-18-38
>54467 MB/s read
>47993 MB/s write
Not bad for the price I guess.
>>55303953
8GB DIMMs, they have higher density ones in super expensive kits though.
What if I told that I cache session data in redis? Will I see no improvement?
>>55299411
>>55299632
Black is entirely gpu bound (8xmsaa)
Red is partially cpu bound, but still very high gpu stress.
>>55303986
Yeah I figured they were 4GB and maybe 8GB DIMMs.
I dont even wanna know how much a 16GB DIMM with those specs would cost.
>>55298672
No reason not to get Z170 because i's newer
Hey, I'm gonna upgrade my ram. It's one of those things I've never worried much about but I've only got 8gb right now so I might as well get 16gb I figure. Who should I go with? I'm not familiar with the companies in terms of ram.
>>55304076
They all use SK Hynix, Micron, or Samsung memory.
Really doesn't matter who you buy from, just choose whatever is cheap with the specs you want.
>>55304095
I've noticed there's four slots in there, should I go with a 2x8 or just pick up two more of the same 2x4 to upgrade?
Not sure how it works for heat dispersion or if those slots are for something else. Again, I really don't know much about ram.
>>55304145
Unless you are on the X99 chipset you want to stick with 2x8GB.
Z170/H170/H110/B150/etc uses dual channel memory so you want 2 sticks, X99 uses quad channel memory (and has 8 total RAM slots) so you'd want 4 sticks.
>>55304165
Gotcha, thanks.
>>55298672
>There is literally no difference in gaming
I think this belongs on /v/.
>or everyday use
Depends on what you do all day every day. What do you do?
>>55298993
Ah, video games. >>>/v/
>implying the price difference between DDR3 and DDR4 is significant
>implying DDR4 isn't a better value for you money
>implying DDR4 isn't the future
>on technology board
>old technology is better
bretty gud bait, :DDD
>>55299271
>not just getting whatever's cheapest and not concerning yourself with what it looks like
