A Proper Phenom II vs i7 Review 5870 Crossfire FINALLY!!!

Raidur

Distinguished
Nov 27, 2008
2,365
0
19,960
Hate to say I told you so...

http://www.legionhardware.com/articles_pages/radeon_hd_5870_crossfire_cpu_scaling_performance_part_1,1.html

&

http://alienbabeltech.com/main/?p=22167

Finally! The results that I knew would show up once we had proper benchmarks done in a resolution lower than 25x16 on a 5870 crossfire setup.

Phenom II is a bottleneck on 5870 crossfire. (Unless at 25x16 res with AA and you OC)

Thoughts, Jenny? :)

In all seriousness, interesting article eh guys? Especially considering how well i3 keeps up (and doesn't), with the Ph x2 not too far behind. Discuss!! (not a flame thread)
 
two things to keep in mind

A conventional LCD monitor runs at 60 Hz . Thats 60 frames a second . Doesnt matter what the graphics card/cpu combo produce or what fraps is telling you you are getting an absolute max of 60 fps actually appearing on the screen in front of you

Second point is value . The phenom costs a lot less
and why no i5 750 in the comparison? Thats a closer price match to the phenom set up
 


This is the first benchmark of this kind that I've seen. It should be news to some people. I know low resolution benchmarks showed us i7's superiority, but a lot of people weren't convinced saying it was a "resolution nobody uses". Now that the GPUs are catching up we're beginning to see the difference, and this is the first bench I've found that shows the difference using every day resolutions with mixed CPU frequencies on GPU's strong enough to relieve some of that 'GPU bottleneck'.

So unless you were one of the "intel is only better at lower resolutions" people *cough Jenny *cough then this shouldn't be news.
 


Totally agree. Smoothness changes for me when I play CS:S on 300FPS compared to 100FPS but I know I'm not actually 'seeing' more than 60, but your point still stands and I agree. :)

I don't know why they didn't include i5, they're doing a Part 2 with c2q and those older cpus, maybe he'll throw in i5. However I'd expect i5 to be within 5% (clock for clock) considering HT / triple channel / 2x 16x 2.0 , don't make much of a difference (yet).
 


Are there other reviews like this one? Ones that show i7 vs Phenom II on a clock for clock basis with 5870 crossfire (or better) properly (matched components) and mixed resolutions? It's the first one I've seen. I knew these were the differences due to low resolution benches I've seen (along with scaling benches), but there are many nay-sayers about the low resolution benches.

Or am I just too little too late? =P
 


It's not that its a bad review or anything its just this has been common knowledge for awhile now. Don't need graphs or a benchmark to know certain things to be true based off of other benchmarks.
 
Okay thanks. I read it looking for surprises. Clock speed for clock speed it's well known i7 beats P II X4 but to see the X4 go to 4 Ghz to just barely keep up with i7 @ 2.6 Ghz was the surprise for me.

Really, if someone is gaming hardcore enough to invest in crossfiring dul 5870's, they're not gonna skimp, save a couple of hundred with a $180 CPU and AM3 mobo. They're gonna spend it on the LGA 1366 platform to push out the highest framerates driven by the fastest hardware.

This article would have been more interesting if the comparison was between the i7 920 and the Core i7-980X Extreme Edition Gulftown and better if they also inluded AMD's 6 core Opterons too.

 
The PhenomII was meant to compete with the Core2Quad, not the i7.

Core2Quad architecture is faster clock per clock than PhenomII architecture, while the i5/i7 architecture is just flat out much better than PhenomII architecture.
People who say otherwise are AMD fanboys/fangirls.
 

+1

PII-955 is already hugely bottlenecking 2xHD4890, let alone HD5870.

proof: i7+2xHD48*7*0 beats PII+2xHD48*9*0

PhII%20vs.%20i7%20HAWX.png


 
I think I remember way back in the day when AMD fans used the minimum FPS as a tool to their side. I think it was when Core 2 first came out..... not sure.

Either way I have said that Core i5/i7 will be less of a bottleneck in the future than a Phenom II will.

I want to see 2xHD5970 CF........ that would be a real test right there.
 
2x5970? Yeah, I'd like to see a test of i7-920 vs. i7-980 and see if there is any bottlenecking on the i7-920 with that setup. Although, lets be real here. Who would EVER play games with 2x5970? Heat+power vs. performance, it just doesn't make sense to me after a point. I can't think of any game that "needs" a single 5970, let alone two.
 


Not for those who would (I am sure there will be some) but more like a guess for the future. If ATI keeps at their current pace, then their HD6K series should be nearly 2x as powerful as the HD5K. So a dual HD5970 setup could be the same as a dual HD6870 setup.

Plus it would seperate the CPUs out since it would need major bandwidth.
 


Then use one one 5970! It's cheaper than 2 5870s. I thought of it because a while ago I remember seeing duals Opterons versus dual Xeons on Crysis.
 


I quote from the conclusion of the very same review:
"The most expensive Phenom II X4 processor is the 965, which costs just under $200 US. The cheapest Core i7 processor is the 920, and at $290 it is considerably more expensive. Then of course you have the associated motherboards and memory kits, all of which are much more expensive for the Core i7 platform.

Something worth considering is that although the Core i7 clearly is the superior processor, can all that extra horse power be converted into better gaming performance? When we conducted our CPU scaling articles with the Radeon HD 5970 we found that the extra power of the Core i7 was of no benefit. This was due to the heavy use of AA/AF, which placed limitations of the GPU’s before the CPU.

Therefore it could be said that when enabling the kind of quality settings that owners of Radeon HD 5970 or Radeon HD 5870 Corssfire configuration would generally play with, there is almost no difference in performance between the Phenom II X4 and Core i7 processors. "
http://www.legionhardware.com/articles_pages/radeon_hd_5870_crossfire_cpu_scaling_performance_part_1,21.html

 
Yeah, he forgot to mention the high resolution (25x16) along with the AA/AF in the 5970 review, that was the major reason.

The main reason I showed this aticle is because it used mixed resolutions with a high end GPU setup to squash the GPU bottleneck. Once we use even stronger GPUs i7 would be in the lead even at the high resolutions and AA/AF. Some people argued against that because before this review we only had benches with low resolutions to relieve the bottleneck, and those were supposedly 'unadmissable'..

Andy:

Yeah I remember that article. I just remember some of the fanboys (jenny, cry, etc) dismissing it for some reason so I hoped this would open some eyes.