It appears that RAM speed does make a big difference in some games!

[TheAudiophile]

I came across these videos from Digital Foundry: 

 

 

 

 

They have a pretty good testing methodology, and it appears that faster RAM can definitely make a big difference in some games. I was under the impression from LTT videos that RAM speed barely made a difference at all?

 

What do you guys think about this? Is LTT's testing just not thorough enough to reveal these results?

[MageTank]

4 minutes ago, TheAudiophile said:

I came across these videos from Digital Foundry: 

 

 

 

 

They have a pretty good testing methodology, and it appears that faster RAM can definitely make a big difference in some games. I was under the impression from LTT videos that RAM speed barely made a difference at all?

 

What do you guys think about this? Is LTT's testing just not thorough enough to reveal these results?

What do I think? I think i covered this subject months ago. I also think OCN covered it years ago. LTT's testing methodology was flawed, big time. They used graphics settings that were completely out of range of that GTX 660 Ti's performance grade, and they also failed to monitor minimum framerates, arguably the most important framerate in gaming, and is most impacted by memory speed. Their quad channel memory speed was slower than my dual channel speed too, but that's a complaint for another day.

 

This is not recent news though. I have been saying this for half a year now, and as i said before, OCN has been saying it for nearly two years now. http://www.overclock.net/t/1487162/an-independent-study-does-the-speed-of-ram-directly-affect-fps-during-high-cpu-overhead-scenarios

[WaitDontSh00t]

I have seen videos like this and this one, I read an article somewhere a while ago, and it said that 2800 is a sweet spot I think, anything above doesn't really provide anymore performance in games at least. I'll have to decide which ram to get today for my new build!

[TheAudiophile]

Well, I guess MageTank's post clears that up

 

LTT should really make a new video on RAM speed, although they do seem to focus more on individual products rather than technical stuff. Maybe that's why they have so many subs? They do seem more accessible to the average PC enthusiast than more technical channels.

[Monarch]

15 minutes ago, TheAudiophile said:

Well, I guess MageTank's post clears that up

 

LTT should really make a new video on RAM speed, although they do seem to focus more on individual products rather than technical stuff. Maybe that's why they have so many subs? They do seem more accessible to the average PC enthusiast than more technical channels.

LTT is all about the show and product advertisement, that's where the money's at. If you want serious tests from people who know what they're doing that's Eurogamer/Digital Foundry, Pcper, guru3d, Anandtech, etc.

[El Diablo]

Digital foundry are badboys

 

linus is doing a completly diferent thing

 

 

[El Diablo]

4 hours ago, MageTank said:

What do I think? I think i covered this subject months ago. I also think OCN covered it years ago. LTT's testing methodology was flawed, big time. They used graphics settings that were completely out of range of that GTX 660 Ti's performance grade, and they also failed to monitor minimum framerates, arguably the most important framerate in gaming, and is most impacted by memory speed. Their quad channel memory speed was slower than my dual channel speed too, but that's a complaint for another day.

 

This is not recent news though. I have been saying this for half a year now, and as i said before, OCN has been saying it for nearly two years now. http://www.overclock.net/t/1487162/an-independent-study-does-the-speed-of-ram-directly-affect-fps-during-high-cpu-overhead-scenarios

Linus and digital foundry are doing two completly diferent things

[MageTank]

25 minutes ago, El Diablo said:

Linus and digital foundry are doing two completly diferent things

I know. One focuses more on entertainment value, the other on providing factual information.