>1600MHz: 36.0/54.6 fps
>2133MHz: 39.0/61.0 fps
>2400MHz: 44.0/66.9 fps
http://wccftech.com/fallout-4-performance-heavily-influenced-by-ram-speed-according-to-report/
For the longest time ram speed was a non issue for the most part
Is this starting to change with new games?
Will X99 and quad channel DDR4 offer bigger benefits outside of content creation?
Or is this just bad game optimization?
>playing shitout 4
I've seen DDR3 up to 2900MHz
Why don't they test with that, surely at some point it shouldn't matter anymore
DDR4 can go way higher too
Guess I'm glad I got cas15 3000 ddr4
>>51319029
>>>/trash/
Seriously this game is an unoptimized piece of shit. Using this game as a benchmark for any sort of reccomendations on hardware is laudable.
So does it only go off of bandwidth?
What about timings and latency?
>>51319088
WITNESSED
I could honestly not care less about 7 frames.
>wccftech
>user defined "benchmark" with many variables to take into consideration
>wccftech
>tech
>wccftech
>wccftech posting anything that isn't an absurd untrustworthy rumor
>wccftech once claimed Kleenix is making high powered GPU's
>wccftech once claimed nVidia was funding ISIS
>>51319125
Can be a big deal desu senpai, essentially the difference between smooth 60fps and stuttery 50fps
>>51319135
>>wccftech once claimed nVidia was funding ISIS
Link?
>>51319029
>Or is this just bad game optimization?
According to that review a Titan X won't get over 70 fps at 1080p maxed settings.
A god damn Titan X.
Sounds like the game is complete shit. I'd avoid it or ask for a refund.
>>51319145
>50 fps
>stuttery
What the fuck
Its a bitch really as not all cpus (primarily vishera ones) can handle fast ram. Plus as a great many /g/ users are retarded they don't know to allow XMP in their bios because most bioses are fucking ancient in design.
>>51319163
>referring to titan x as a flagship
xd
>>51319166
geez you must have low standards, console peasant.
>>51319088
>>>>/trash/
>not just a redirect to [s4s]
>>51319174
Woah there, do you know how fast you were memeing back there?
>>51319166
ya for example try gta on 50 fps, very stuttery compared to 60fps
>>51319194
>doesn't understand microstutter
>>51319125
>http://wccftech.com/fallout-4-performance-heavily-influenced-by-ram-speed-according-to-report/
you're a faggot, if you were getting 23 FPS you would care about 7 more frames. it's all relative. 7 frames is a lot no matter how you try to push it, you could be 7 frames from the annoying laggy boundary of 30 fps, or you could be 7 frames higher than 60 fps(aka 67), giving you a nice buffer if you encounter a graphically intense area that would otherwise cause frames to crash. having a high fps buffer means there's less room for sudden changes in graphic intensity to plummet visual performance to subpar levels and can also be used as a means to keep your framerate above a set refresh rate to prevent tearing or input lag.
gtfo.
>>51319173
It isn't? Then what is? The 980 ti, which is just a gimped Titan X?
>>51319253
*i should say there's more room for sudden changes in rendering intensies to NOT plummet frame performance to subpar/unpleasant levels
>>51319194
>>51319174
I can feel the difference between 40 fps and 60 fps when I play battlefield on my laptop, but i still wouldn't call that lag, it's still very playable.
Obviously it's more enjoyably at 60 fps, but come on, 50 fps means laggy? That's just placebo.
>>51319256
the 980ti is better for video games
(and it still runs fallout 4 terribly)
>>51319315
mine does fine
Fuck all this noise I'll just play on ps4
>>51319029
>970
>texture swap on 0.5 to system ram
>ram speed is influential
no shit sherlock
>>51319029
>>51319348
IIRC the 0.5 is still substantially faster than system memory
>>51319291
You need 30 fps for each eye. You're LITERALLY fucking placebo you mustard fart
>>51319029
>Or is this just bad game optimization?
>bethesda
>>51319348
fallout 4 doesn't use 3.5gb of vram
>>51319392
You only need 24 FPS per eye, ya moron. Unless you wear glasses, then you need 48 FPS per eye, since each eye counts as two now.
>>51319424
they don't even state the resolution, for what we know that shit could be 4k
>>51319390
It isn't. but it's faster than pci express 3.0, which is the bottleneck for swaps.
>>51319392
anything less than constant 60 is a pain in the ass, especially in games where input is affected by fluctuating fps like fallout 4 is.
>>51319029
so much gains from memory alone is really shitty code from bethesda.
>>51319438
At 3200x1800 Hwinfo64 was reporting 3-3.3gb of vram usage for the 20 mins or so I played at the resolution on a 290x.
>bought a new build
>got 2800mhz RAM
Feels fucking good.
Are all i7s better than i5s?
> buying ddr4 in its present state.
Yeah I'll wait.
>>51319088
>laudable
I guess you mean laughable.
Someone post the graph that shows the 1 frame difference between previous RAM speeds.
Is Fallout 4 the new Crysis?
>>51319522
It is no where near as good looking as Crysis.
skylake haters BTFO
This was also observed in GTAV.
Making an (un?)educated guess, this has something to do with the games being ports of the console version and their different handling of RAM etc.
>>51319510
For most games ram difference is nonexistant, but if bethesda cannot into memory properly ram speed can matter. A good follow up to these claims is a latency test as well.
>>51319486
Pretty sure i5 6600 would beat an i7 860
>>51319534
You mean keklake users btfo.
You don't got that sweet sweet quad channel RAM with x99.
>>51319059
Only a dubs thread if you GET dubs
If I have 1600MHz RAM but it has relatively low timings, could I reliably boost it up to like 2400 and just increase the timings?
Is this actually beneficial at all? What's a program I can actually witness the benefits of faster RAM in. I hate people who just post game benchmarks of negligible framerate changes to show RAM overclocks.
>>51319566
Nice GET; now a dubs thread.
>>51319029
>this game is such a piece of shit it doesn't even use VRAM and stores everything in RAM instead
sasuga bethesda
>>51319588
RAM is less expensive than VRAM though
>>51319566
>>51319577
>>51319588
This thread destined to be a dub thread
>>51319599
Useless observation
>>51319612
>yfw 77 88 99 consecutive dubs
>>51319566
>>51319577
>>51319588
>>51319599
oh boy
>>51319315
980ti here. 1600 DDR3 16GB. SSD game install. i5 4690k
60-70 FPS on ultra with a couple mods.
Sorry your PC sucks nuts.
You might need this book
>>51319174
/v/ go home.
>>51319029
Why the hell is it dependent on frequency and not latency like everything else out there?
the human eye can't see above 24 fps anyway. why are we still discussing this?
>>51319315
gaymes ain't the only thing a GPU is for.
The Titan X is Nvidia's flagship GPU. The 980 ti is the best gaming GPU they have (according to you but I'll take your word for it).
Quadro M6000 is the real money maker tho.
>>51319166
50 FPS is trash. If you are incapable of building a decent PC, just kill yourself. You're obviously poor or retarded
>>51319253
>>51319322
Peasant
>>51319481
>bought a new build
Pre-built detected
How's your HP memeputer?
>>51319673
I play on my laptop battlefield 4 35-45 fps perfectly fine during lan parties.
>>51319671
>using a titan x for anything but gaming
>>51319651
What's this from?
>>51319777
Digital Foundry.
>>51319777
are you a moron ?
WCCF is just reporting on some one elses benchmarks/findings.
http://www.eurogamer.net/articles/digitalfoundry-2015-the-best-pc-hardware-for-fallout-4-4023
They link them in the article.
>always go to the source
>>51319777
The internet
>>51319801
This is just lazy benchmarking. They put all the settings on Ultra, when some of the Ultra settings are objectively bad.
God Rays -> Low
Shadow Quality -> Medium
FPS -> Doubles from "everything on Ultra"
http://www.logicalincrements.com/games/fallout4/
>>51319844
>my card can't max this game therefore nobody should use max settings as a benchmark
AMD logic
>>51319877
A 980ti struggles.
>>51319898
i have a bulldozer cpu and a meme gpu and can play just fine with ultra settings, their setup was bad
Is dual-channel a meme? Every gaming benchmark I've seen shows it doesn't even do anything.
>>51319844
why do the rocks and grass change when going between the shadow quality?
>>51319936
no, it just doesn't benefit gaymes as all of them are poorly designed OOP garbage.
anything that does lots of bulk memory operations (compression, encoding, etc) will benefit a lot from it
>>51319029
It's funny how pathetically ugly the graphics in this game look compared with how demanding it is. Games in 2005 looked better than this. Clear proof how actual design of the graphics is more important than shader effects and polygon count.
>>51319944
Small rocks and grass are not static. They are dynamically generated per load.
At least, that is what I saw when I kept reloading and comparing screenshots.
>>51319844
Did people complain this much when crisis came out and could barely be played on an 8800gt on ultra?
Who cares, it doesn't run well on a 980 ti? It'll run fine on a 1060 ultra settings.
Just you wait. It'll even be playable on Intel 7th gen i series igpu
>>51319936
>Tfw quad channel
>>51319955
Only when working on a server size scale.
http://www.gamersnexus.net/guides/1349-ram-how-dual-channel-works-vs-single-channel
>>51319985
Is this a sarcasm or a bait? Im retarded and cant tell, please halp.
>>51319936
Dual channel was a good solution to a real problem back in the day: RAM bandwidth bottlenecks.
Ever since DDR3, RAM has not had any bandwidth bottlenecks, so it is a solution to a problem that no longer exists.
>>51319985
I get 50-60 FPS on Ultra (with God rays low/off) on a 290.
On 1440p.
People complain because their powerful computers are getting low performance on a non-ultra-good-looking game.
Bethesda should not have included God Rays in the stuff that changes with settings. It should have been an independent option, like resolution or AA.
>>51320018
I know this is poorly optimised and crisis was good, but my point is, if you don't like the game don't buy it. If you're going to buy it you know what you're getting into, complaining about it won't make a difference.
Current hardware might not be powerful enough for such a shitty game but next gen will be. It's a game. You won't die if you don't play it.
>>51320058
i'm not even concerned about fps, things seem smooth enough for me and I just want to enjoy the game.
>>51320070
>Current hardware might not be powerful enough for such a shitty game but next gen will be.
But it is!
>>51320096
If it is then why are so many people bitching about poor performance?
>>51319029
>an unoptimized game runs like shit on slower hardware
maxed settings on a 970 when loading a new area
>>51320058
really its not gpu problems that are hurting performance - its whatever the fuck bethesda is doing to the cpu. People with far more powerful cpus even my 9590 are reporting horrendous fps drops while their gpu sits below 100% usage. Whatever bethesda are doing its molesting cpus for no reason.
My own limited tested showed that npc corpses causes horrendous frame dips. Looking out over the game world with max distance? Fine. 5 dead supermutants? lol frame drops.
>>51320103
Because Bethesda's current settings are bad.
>>51319083
I think the highest specification is 6400
Otherwise DDR4 is nothing different
My 2133 DDR3 should keep me up for a while
>>51319163
Possibly, my friends are hyped, but I think its not worth it
I own a Titan X and would possibly need to play it on medium @4K
>>51319936
It is, mostly
For me, switching from 1600 to 2133 RAM reduced microshuttering in Mass Effect 3 Multiplayer. so it was worth it. It should work also in better optimized games.
>>51319088
>laudable
>>51320130
Elaborate.
>>51319029
lol what you have 50 fps on that rig? i get like a hundred or so on my 2500k 280x
>>51320193
They should have grouped God Rays with resolution/AA, instead of tying it to presets.
>>51320230
No, it shouldn't be grouped with anything.
There should be a 'pleb' and 'advanced' mode, and in advanced almost nothing should be tied together.
Is new vegas just as fun as 4? i was going to have a go at that instead of buying the new one
>>51320772
they're both good. 4 is more modern and more of an action game and NV is a bit clunkier and more of an RPG. i'd say 4 is more easy fun but NV has a better payoff if you like good, well made RPGs
>>51319654
>intel i5 4690k
So you have a poor person processor that is a year old? Nice meme my newfriend
Had to scrape by to grab that gpu?
:^]
>>51321042
Not really poor, AMD is poorfag tier. If I had no money I would choose a 2500K OC for cheap.
>>51320111
Thats caused by a pirated version you idiot, I have a legit copy and never had this problem
>>51319029
>8fps
Holy shit that's a lot!
>>51319211
well you stupid nigger its not really about ur eye its about input lag, which is very noticeable at those respective fps'
So is this the first game that benefits from faster RAM frequency?
>>51319407
ya that's the point desu senpai
>Their i5 4690k & TITAN X with 1600MHz RAM getting 36/54.6 fps
>Tfw my 4690k & gtx 980 with 1600 ram gets a locked, solid 60 WITH dsr activated
>>51321360
That's over a 15% performance increase, that is absurd
>5930K
>980 ti
>3333mhz ram
Too bad bethesda games are more boring than watching paint dry
>>51320111
posts pirate copy blames Bethesda
HAHAHA
>>51319166
Anything that's not an integer division of your refresh rate will look stuttery.
e.g. if you have a 60 hz monitor, if you run the game at 59 FPS then every second, a frame will have to be repeated which makes it look stuttery.
>>51319898
None quality? Why is it not infinite fps?
>>51319029
>those graphics
>"ultra"
>2015
nice going bethesda
Check mt single
>>51319029
It has nothing to do with the age of the game and everything to do with the amount of people called "rajeesh" in the programming team.
>>51319029
>Is this starting to change with new games?
no. this game is a broken piece of shit. it's not a benchmark for anything except a disaster in software engineering.
>>51319059
>>51319088
>ram isn't technology
>>51319898
holy shit, that poor 980TI, the game doesnt even really look that great to warrant this. Stalker cop pretty much looks better than this game and its already god known how old.
The TI should be able to run twice, 3 times as fast at least on that resolution.
>>51319029
>Skyrim graphics
>DDR5
>>51319444
Fallout 4 actually runs faster or slower depending on FPS. So too much FPS for that game is actually also a bad thing.
Seriously, go check out some videos. I don't know how the fuck that game made it to release in the state it is, but I'm glad I'm not a Fallout fanboy.
>>51319936
it doubles memory bandwidth. Anything that is bottlenecked by memory speeds will see a huge increase in performance. People try to boil this down to a % increase in general, but it really is application specific.
>>51319029
WHAT IS THAT SHIT
OMFG LMFAO
MY SIDES
>>51319612
>>51319637
let's get em boys
![1428161466380s[1].jpg](https://i.imgur.com/BYSEeMa.jpg "1428161466380s[1].jpg")
>tfw 40-50 fps with gtx 970
>zero motion blur or smoothing so anything below 60 looks like stuttering shit
>turning down the quality only slightly improves fps
>I spent $350 for this
p-p-c master race, r-r-right guys?
>>51319029
>Is this starting to change with new games?
considering they also locked the physics to the framerate, I'd say no. bethesda is just incompetent.
>>51326498
Fallout fans hate Shitout 4 the most, senpai.
>>51319029
thankfully I just ordered some DDR4 2800mhz for my new rig;)
>>51326888
Check'd
>>51319566
checked
Check this 5.
>>51327727
lmao
>>51327727
tbh fam
>>51327727
that can't be real
>>51319054
thats not the point dingus.
the point is the actual increase in performance.
i have 2.4ghz ram so this is interesting.
>>51320164
really? you actually noticed a difference? dang man, i have never noticed anything different
because 1600 can run at cas 9-9-9-15
while 2133 runs around 11/12 which should actually make the delay even larger.
>>51319898
Something wrong. 8320 and a GTX 960 here and maxed out at 1080p I get 55-60 indoors and less demanding areas and in big fights I go down to 40-45. 980 ti Should get 60 easy.
>>51327727
What am I looking at here?
>>51328039
At long last, Hiro has stopped the filtering.
We can enjoy our tbhs and senpaitachi in peace!
>>51319054
This is running counter to 99% of 3d rendered applications which depend on memory latency for a 2-3 fps boost over raw frequency which can actually drop frames, if you string out your latency to get it.
The only possible explanation I can see is that something is fucked up and isn't working fully or they did something relating to memory very right and can finally make use of faster speeds.
>>51328241
https://www.youtube.com/watch?v=Q894F3TPXeU
console performance, genius
>>51321158
>Thats caused by a pirated version
How so? Not heard of this yet
>>51319651
>970
>0.6% dropped frames at 1080p
>11.3% dropped frames at 1440p
You can't make this shit up
>>51319029
This is likely because of how newgen consoles have a unified (very fast) memory which both the cpu and gpu uses. Because of this they probably do certain operations in a specific way which accounts for it, also memory is speed less of an issue for them. On the other hand on PC while vram is fast, only the gpu ueses it (mostly). All the other shit resides in the system ram, which is not so fast, yet memory is more or less probably managed the same way.
>>51328341
So Bethesda were too lazy to do a proper port.
>>51328260
Holy shit that's fucking disgusting
>>51326888
Gud tribs
>>51328341
So its a hardware issue? How come other open world games don't have such big issues?
>>51328260
This shows massive fps drops in tght corridors as well. This is just a shitty designed game, nothing more.
Only a piece of shit company like Bethesda can make a game that looks shit AND runs shit.
>>51319029
>For the longest time ram speed was a non issue for the most part
>he literally fell for this meme
>>51328321
>performed better than comparative card
Why is that funny?
> Fallout 4 turned out to be an unoptimized piece of shit
Who would have guessed.
> but le modders will fix this :)))))
Just fucking hang yourself.
>>51328386
>So its a hardware issue? How come other open world games don't have such big issues?
Not really. It's the issue of being differences between a console and a pc in certain ways (less differences than with older gen consoles) and the developer didn't account for this properly. Probably as a way to try to cut down development costs: they wanted to rewrite less shit.
>>51328321
Both are lower than the 390X
>>51319029
>Or is this just bad game optimization?
This. Shitesta fucking sucks dick.
>>51321939
>>51321158
>mfw bought copy at the store
>mfw those concrete walls
Not to mention you fall through them
GG Bethesda
>>51319898
Lel.
Turning God rays to low really does have a massive impact. I'm getting 60+ with only occasional dips to the lower 50's at 2k with a Fury X with God rays low, and 16x TSAA.
This game is broken.
>>51320111
>Metal Gear Solid
>Ps1
>1998
>>51319166
anything less than 60 fps is trash
