[DigitalFoundry] Is It Time To Upgrade Your Core i5 2500K? - Page 29 - Overclock.net

Forum Jump: 
 
Thread Tools
post #281 of 402 Old 02-28-2016, 03:41 AM
M'aiq the Liar
 
Darkwizzie's Avatar
 
Join Date: Jul 2012
Location: Bay Area, California, USA
Posts: 7,383
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 422
Quote:
Originally Posted by Partol View Post

Are there any good game benchmarks which would show a significant difference by changing the ram clock/timings?

I just tested the Mafia 2 in-game benchmark: Once with my normal ram settings and once again with ram timings doubled (but same ram frequency).
The difference was less than 1 fps.

People say that fallout 4 is very ram dependent but is there a benchmark for it?

Instead of a benchmark, I suppose I could create a save game at a starting point, and then just move forward and record it. And retest from the same starting point.

I think Fallout 4 is CPU dependent. With my 6600k and 980ti, the GPU is usually at 100% or near 100% usage with frame rate uncapped. The times when it falls below 60 fps however, are usually due to GPU usage dropping. And the times when it falls below 45 are basically exclusively GPU with low usage, CPU going crazy. Some of what I have seen suggests that when it's not GPU dependent, it's really often CPU + ram dependent. I have not tested this myself. I guess I could, though.

 

So when we look at Fallout 4 with an underpowered GPU with god rays turned up, or we just benchmark a typical area in Fallout 4, we only get part of the picture. The areas in downtown Boston where frame rates take a hit are not the majority of the game, but when the fps does drop it can be a little jarring. And so, this is why I think the CPU and GPU testing with Fallout 4 is incomplete. It doesn't account for the problem areas of the game.

 

Although, I haven't retested with the new patches.


My Overclocking Charts (Click to show)Computer History (Click to show)
??? - Some computer long long ago, far far away
2001 - Random Windows 95 computer that sucked
2002? - Emachines - Tualtin CPU, 1.3ghz, 256mb SDram, HDD 40gb (Unnamed)

2008 - Wizzie Machine - Q6600, 2gb DDR2, SLI 8800GT 512mb, HDD 'Wizzie Hub' 250gb HDD

2013 - Celapaleis - 4670k, 8gb DDR3, 7970ghz 3gb, SSD 'Celapaleis', 256gb 840 Pro
2015 - Celapaleis (Reprise) - 6600k, 16gb DDR4, 980ti, 840 Pro, 850 Pro, Pack Yak II
2017 - Undelwalt - 7600k, 16gb DDR4, 1080ti, 950 Pro, Pack Yak II

M'aiq wishes you well.

Darkwizzie is offline  
Sponsored Links
Advertisement
 
post #282 of 402 Old 02-28-2016, 05:38 AM
 
BigMack70's Avatar
 
Join Date: Jan 2012
Posts: 3,227
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 184
Quote:
Originally Posted by Majin SSJ Eric View Post

Even a Titan X won't be held back much by a 2500k at something like 4k resolution. Sure, at 1080p you will likely find some CPU bottlenecking (depending on what GPU(s) you are running) but really, if you are willing to spend $1k on a single video card you should probably not be running a 5 year old platform to begin with...

Can't comment on a 2500k but with a 2600k, the only games where it held me back at T-X SLI at 4k resolution were GTA V and a little bit in Crysis 3. Crysis 3 was still able to get ~60fps but that bumped up about 15-20% with a 5930K. GTA V didn't see framerate improvement but did have some stuttering when maxed out on a 2600k that was gone on a 5930k.

No change in my other games. Granted that's a 2600k and it was at 4.8 GHz but at high resolution you're just so much more GPU bound than CPU that the CPU is really not a very important piece of the puzzle most of the time.

BigMack70 is offline  
post #283 of 402 Old 02-28-2016, 10:24 AM
1.3ghz
 
Partol's Avatar
 
Join Date: Feb 2011
Posts: 734
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 34
Thanks for the suggestions/encouragement. The built-in mafia2 in-game benchmark seems worthless for benching ram. Maybe it pre-loads everything into ram.

Fortunately, in mafia 2 - chapter 2, there is a scene in which the car autodrives itself through the city. Excellent scene for doing comparative benchmarks. The results were more that I expected.

http://www.overclock.net/t/1487162/an-independent-study-does-the-speed-of-ram-directly-affect-fps-during-high-cpu-overhead-scenarios/190#post_24941318
Cyro999 likes this.

Partol is offline  
Sponsored Links
Advertisement
 
post #284 of 402 Old 02-28-2016, 10:38 AM
 
santi2104's Avatar
 
Join Date: Apr 2014
Location: Buenos Aires, Argentina.
Posts: 144
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 6
I have been using the sandy bridge platform scince 2012, and i cant find an excuse to upgrade, for gaming it still kicks ass, 4.4ghz stock vcore rock solid for years, long live sandy bridge

santi2104 is offline  
post #285 of 402 Old 02-28-2016, 10:44 AM
 
madpossum's Avatar
 
Join Date: Dec 2014
Posts: 364
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 67
Quote:
Originally Posted by Majin SSJ Eric View Post

Even a Titan X won't be held back much by a 2500k at something like 4k resolution. Sure, at 1080p you will likely find some CPU bottlenecking (depending on what GPU(s) you are running) but really, if you are willing to spend $1k on a single video card you should probably not be running a 5 year old platform to begin with...

Your going to see a lot more difference upgrading your GPU than CPU in almost all cases. If they don't upgrade very often and then find themselves with around $1000 to upgrade I could see them putting all that into the GPU if they are still happy with their CPU, instead of upgrading both and ending up with a slightly better CPU and a 970 or something. Though I personally would have worked the numbers or saved a little more and upgraded the CPU and get a 980 Ti (which is what I did).

I have come to chew bubblegum and kick ass.....and I'm all outta bubblegum.
madpossum is offline  
post #286 of 402 Old 02-28-2016, 10:56 AM
Fan Man
 
Smanci's Avatar
 
Join Date: Jan 2012
Posts: 1,670
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 76
I just swapped my 3570K to 3770 for basically no cost. It does help in some newer games and also allows 41x multiplier + BLCK bump if I ever feel need to OC.

Smanci is offline  
post #287 of 402 Old 02-28-2016, 12:18 PM
Case Modder
 
Cyb3r's Avatar
 
Join Date: May 2012
Posts: 339
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
honestly the problem with fallout 4 and skyrim is that they feature areas where no matter what hardware you throw at them you'll still see low fps (waterfall in the first town in skyrim for example : while i did notice a difference between my 680 and my 980ti with the exact same settings and making sure i didn't overrun the vram on the 680 (which is rather easy to do with mods and the 680 since it only has 2gb off vram) the difference boiled down to about 3-10'ish fps in that area since it's a known problem with the actuall engine

Cyb3r is offline  
post #288 of 402 Old 02-28-2016, 12:41 PM
1.3ghz
 
ASUSfreak's Avatar
 
Join Date: Nov 2008
Location: Belgium
Posts: 2,511
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 135
Quote:
Originally Posted by Darkwizzie View Post

With my 6600k and 980ti, the GPU is usually at 100% or near 100% usage with frame rate uncapped. The times when it falls below 60 fps however, are usually due to GPU usage dropping. And the times when it falls below 45 are basically exclusively GPU with low usage, CPU going crazy. Some of what I have seen suggests that when it's not GPU dependent, it's really often CPU + ram dependent. I have not tested this myself. I guess I could, though.

Wow, I posted something on page 5, now it's like 28 or so tongue.gif

I had Core 2 Quad Q9450 once with SLI GTX470's.

When I swapped it for my 2600K and the same 470's they got HUGE headroom and my fps went UP!

Ofcourse my Q9450 was bottlenecking my 2 470's. The cards produced much more data per second than my proc+chipset+ram could follow smile.gif so that was bottleneck.

With upgrading to P67 and 2600K the fps doubled or something smile.gif

But when I bought 1440p monitor they could not cope anymore so bought 780Ti's

Greetz ASUSfreak helpinghand.gif (PS: I made the Avatar myself --> the "freak" part cool.gif)

The first ASUS logo of the nineties...
http://www.speedtest.net/my-result/4988346479 --> 200/20Mbps
http://valid.canardpc.com/show_oc.php?id=1866660 --> 5.2GHz
http://www.techpowerup.com/gpuz/wxudh/ --> GTX470 930MHz/1000MHz/1860MHz
http://www.techpowerup.com/gpuz/details.php?id=z5hq7 --> GTX780Ti 954MHz/1800MHz/1174MHz
ASUSfreak is offline  
post #289 of 402 Old 02-28-2016, 12:47 PM
Apple Enthusiast
 
TMatzelle60's Avatar
 
Join Date: Aug 2014
Location: Levittown, NY
Posts: 1,595
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 16
I believe the 2500K is just like the 460 was to Nvidia. Its a well made beast i think a lot of people will have it for 3-4 years more.
TMatzelle60 is offline  
post #290 of 402 Old 02-28-2016, 05:08 PM
News Fiend
 
jetplane48's Avatar
 
Join Date: Dec 2009
Location: SoCal :]
Posts: 2,189
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 48
Maybe it's time for a lil change smile.gif

teaching.gif
What if air is actually poisonous and it actually takes ~80 years to kill us?
What if water is an addictive drug and the withdrawals kill us a few days later?
What if we are close to death and we are living life in a flash at this moment?
teaching.gif
jetplane48 is offline  
Closed Thread

Quick Reply

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off