Gaming on Intel or AMD...The truth

Printable View

Show 40 post(s) from this thread on one page

06-06-2013, 02:20 PM
The Joker

Quote:

Originally Posted by Wesley

Stick a FX-6300/FX-6100 in there and that would be an interesting test. No Crossfire or SLI, though. We wouldn't want Nvidia's frame metering tech to smooth out the frame delivery while Xfire with VSync disabled is almost a complete mess.

I'll try organize a FX-6300, as for sli and cf, I am not gonna bother with that for now.
Most normal gamers don't bother with multi gpu setups.
06-06-2013, 02:26 PM
Isengard

Quote:

Originally Posted by The Joker

http://www.anandtech.com/show/6985/c...-in-haswell-/5

The above mentioned article is the reason I am doing the test. Some people still claim that intel is better for gaming bla bla.

Well this is how its going down.
The cpu's will be run at stock clocks with HT and all cores/modules active.
Each machine will be running the same memory (G.skill DDRIII 2400 16GB at its rated speed)
They will be running the same SSD's (OCZ Vector 128GB)
I will not be showing any fps numbers during the test, I want it to be a "blind" test

If you guys can think of anything that needs to be added or changed please let me know.

Are you going to downclock the AMD's to give the Intels a fair shot? :p
06-06-2013, 02:37 PM
Wesley

Quote:

Originally Posted by jasong

Truthfully there is quite a difference in gaming between AMD and Intel at the high end.

Quote:

Originally Posted by The Joker

Some people still claim that intel is better for gaming bla bla.

There is quite a marked difference in frame time delivery between the platforms. If you go way, way back to Tech Report's original story on frame time latency, they had a large test with 18 CPUs and they all showed frame variances.

Inside the second: Gaming performance with today's CPUs

Joker, you're correct about the bottlenecking that you see in the Anandtech graph and that even the A10-5600K seems to run pretty well with the other chips, but they're only noting average frame times. Parts of games like Last Light, Civilisation V (actually, lets just say most of Civ V) and DiRT are GPU-dependent, whereas other parts, perhaps even an overwhelming majority, is CPU-dependent.

http://techreport.com/r.x/cpu-gaming...-beyond-50.gif

By only noting average FPS instead of the minimum and maximum frame rates and the way they oscillate, those graphs mean bugger-all. I like what you're doing, but I wish Anandtech would just move to FCAT benchmarking already.

Once I've got my benchmarking rig set up, I can switch to using FRAPS data for my reviews. I did some testing a while ago and although my Athlon X3 and HD6870 keeps up well, in the graphs I can clearly see where I was experiencing stutter and for how long. Seeing those results for the first time gave me data to use to smooth out my experience by adjusting settings until I had less frame variance.

Quote:

Originally Posted by The Joker

I'll try organize a FX-6300, as for sli and cf, I am not gonna bother with that for now.
Most normal gamers don't bother with multi gpu setups.

Or just disable one module and clock the other FX-8350 down ;-)
06-06-2013, 07:57 PM
simon

I don't see how it could make a difference. If you get 50fps on one machine and 50fps on the other it is all the same. The only thing I can think of that makes a difference is AMD's tend to be slightly louder.
06-06-2013, 08:34 PM
Glordit

Quote:

Originally Posted by jasong

Is that really a good idea? Then all you're proving is that most gamers don't notice a difference between 40 and 50fps.

http://www.anandtech.com/bench/Product/697?vs=701

Truthfully there is quite a difference in gaming between AMD and Intel at the high end.

Yes it's the point exactly, if you don't know the fps count of the game yet it plays the same on 2 PC's with different hardware at the same settings, what's the point of calling one better than the other.
07-06-2013, 07:35 AM
jasong

Quote:

Originally Posted by Glordit

Yes it's the point exactly, if you don't know the fps count of the game yet it plays the same on 2 PC's with different hardware at the same settings, what's the point of calling one better than the other.

10fps might not be that perceptible at high rates but if you end up in a scenario where you're oscillating between 30-40fps from one CPU to the next you will definitely notice it. And after some time your average fps on any rig are going too come down to that level eventually - Crysis3 will strain any GPU on the market.
07-06-2013, 11:57 AM
rootworm41

Intel is better for gaming in the long run, if you are not going to upgrade often. The newer games will start to utilize more cores though so AMD sometimes catches up, especially when you are streaming a game.
13-06-2013, 07:33 AM
Lothy

Quote:

Originally Posted by The Joker

Ok so as you guys know I am into the hardware scene. Love everything tech related :love:
But for a few years now I've been getting pissed off with people claiming they can tell the difference between Intel and AMD while gaming and that Intel is def better. Now I am sorry but I own a i7 3930K, i7 3770K and 2x AMD FX8350 machines. I game on all of them and I usually share cards between them (7970 Matrix, 680, Titan, 7870 cf, 660ti sli). I can not tell the difference between them. They only time I see a difference is when I benchmark or while running fraps.

I was at a AMD conference a few years back and they had some of the top guys from AMD talking about their hardware. According to them CPUs are not as important today as what they use to be. Games rely more on GPUs and only a fraction on CPUs, the only games today that i still see relying heavily on CPUs is MMOs. In my own personal experiments, I had a duo core intel but running a AMD 5970 and i was able to run the game at full gfx and during that time i monitored CPU usage which was nominal at best, most of the performance came from the GPU.

So its it viable still to fork out the money for a new intel (which always seems to come with a new socket) imo no. As long as u have 8gb memory and a good graphics u will game away without noticing anything

The machines will be running the exact same hardware, only the board and cpu will differ.
Games will be set up to run the exact same settings.

So my plan is to do a blind game play test on some mates and possibly some of you guys if your keen :)

The idea is to let my mates play on each machine, machines will be hidden and they won't be able to see system specs.
After letting them play on the 4 machines they then have to tell me which machine they were playing on (Intel or Amd). I want a fairly large group, ranging from seasoned hardware nuts like myself to people that really don't care what they play on.

What do you guys think?

Got to love it when u type up a msg and it doesnt go through...

In a nutshell, gaming is more reliant on GPU these days then CPU. Only games iv noticed where you see a noticeably difference is MMOs, they seems to rely more heavily on CPU
13-06-2013, 11:19 AM
VegOtter

That is an awesome idea. I also don't really see how you can notice gaming performance differences between an Intel and AMD. I would say maybe when it comes to CPU intensive apps like Adobe Premier or aftereffect, then maybe... with a big maybe.
13-06-2013, 04:13 PM
Isengard

Quote:

Originally Posted by VegOtter

That is an awesome idea. I also don't really see how you can notice gaming performance differences between an Intel and AMD. I would say maybe when it comes to CPU intensive apps like Adobe Premier or aftereffect, then maybe... with a big maybe.

There are some games that I know for a fact you'll be able to tell. It's in the minority though.

Show 40 post(s) from this thread on one page