Sigh.....I really didn't want to have to answer this thread but now I have to.
wizardsmagic007 Wrote:I got my processor and motherboard for $150 total. So in my case Intel was a more expensive route.
Well that's very unusual.
On newegg an fx 8120 goes for $160. The cheapest AM3+ motherboards are around $50 and the good AM3+ motherboards are $100-200 depending on the board. This is about the cheapest you can get them for anywhere. So you must have had a massive discount (which means your retailer was desperate to get rid of them).
wizardsmagic007 Wrote:How much faster does this game run on Intel?
"Intel" is not a cpu. Intel is a company that makes many different cpus.
wizardsmagic007 Wrote:I'm talking about a 6-man match, there's no videos around except the one I uploaded. So don't tell me you know when there's no proof. Go to youtube and see how poorly Mario Party 9 runs on a i5 2320 while my FX-8120 runs it full speed.
We really shouldn't even need to tell you why youtube videos aren't a reliable source of benchmarking information. There are too many reasons to count.
You could beat an i5 2320 running at stock clocks if you're running your cpu at 4.5GHz (and even then not by very much). But that's not a fair comparison. If both cpus are running at stock clocks the 2320 will win. And if both cpus are overclocked with comparable cooling the 2320 will win.
wizardsmagic007 Wrote:Also I upgraded from a Phenom II x4 and the FX is much faster and overclocks much better.
In most benchmarks with 4 or less major threads the FX 8120 performs similarly to a phenom II 955.
The FX 8120 does OC very well compared to a phenom II but only because it's artificially clocked much lower at stock than its brother, the FX 8150. A phenom II 955 can be OC to about 4.2GHz on air cooling while FX 8120 can hit 4.5-4.8GHz and the FX 8150 can hit 4.8GHz. However the FX series have higher temperatures and power consumption at those clock rates. We should also point out that the FX series cpus support a dynamic overclocking system at stock called turbo core. The clock rate range at stock for the 8120 is 3.1-4.0 GHz while the FX 8150 is 3.6-4.2 GHz. This means that if you compare the lowest stock clocks the 955 can OC by about 20% while the FX 8120 can OC 50-60% and the FX 8150 can OC 33%. However if you compare their maximum stock clocks (which they should either be able to reach or at least be able to get pretty close to when running dolphin) the FX 8120 can OC by 12.5-20% and the FX 8150 can OC by 14%. As you can see their OC ability (expressed as a percentage) are actually pretty comparable to phenom II after you factor in turbocore.
wizardsmagic007 Wrote:See, that was my point, it also stands for WWE 12 in a 6-man match. People sit here telling me that even an i3 can run a 6 man match full speed but no one has proved it.
Nobody has said that. All we've stated is that at stock clocks the i3 is faster with dolphin in general, which is true.
wizardsmagic007 Wrote:I already know the outcome, Dolphin is obsolete, so the results will be within margin from a variety of different hardware.
What? That sentence didn't make any sense.
wizardsmagic007 Wrote:Brave for example runs no better on an i5 than my FX.
Which i5? On an i5-2500K it definitely will.
wizardsmagic007 Wrote:It's just funny hearing everyone going on how Intel is so much better for Dolphin
Intel is not better for dolphin. Sandy bridge is better for dolphin. Sandy bridge is a microarchitecture developed by Intel.
wizardsmagic007 Wrote:but in my case my FX can handle any game just as good as an i5 or i7
Which i5 or i7 and what clock rates are you comparing them at? That statement is far too broad to make. An example of where it would be wrong is the fact that an OC FX series cpu cannot beat an OC i5 or i7 desktop sandy bridge cpu in dolphin.
wizardsmagic007 Wrote:It's also funny that my friend has a x4 955 oc to 4ghz and my FX at 4ghz runs games faster with the same Dolphin configuration.
I highly doubt this. The IPC of bulldozer is higher than K10/K10.5 in the vast majority of benchmarks with 4 or less threads. Make sure that you are both using the same build, the same settings, the same game, the same spot in that game, and that you are both not running any background applications.
wizardsmagic007 Wrote:I'm just wondering if anyone actually did any testing or just believe what they read online?
We believe the data we see online. There are countless individuals and organizations that have benchmarked FX series of cpus with a wide variety of applications and put their results up online for all to see. Are you implying that they are all wrong? How would thousands or organizations all get the same incorrect results?
wizardsmagic007 Wrote:I also have a few Core 2 Quads and that cpu hierarchy chart for dolphin is bogus. Core 2 Quads can't compete with FX's they should not be in the same category.
Yes they can. They are comparable to high end core 2 quad cpus like the 9650 in applications that use 4 or less major threads both with and without OCing. Although I do admit that on second thought the FX series cpus should be moved from category 4 to 3 since they are faster than most core 2 cpus and comparable to the top of the line core 2 cpus. FX series cpus can match the speed of category 2 cpus if they are OCed but not category 1 cpus.
The reason my thread is inaccurate is because their are many different models in each cpu family. I can't list them all because their are too many. And since the difference in performance is so great going from the lowest model to the highest model in each series there are always going to be some instances where it is wrong. Just look at core 2 duo for example. Core 2 duo cpus range from 1.6 to 3.4GHz, from 2MB of L2 cache to 6MB of L2 cache, and from 800 MHz FSB to 1333 MHz FSB. The top of the line core 2 duos are 2.5x as fast as the bottom of the line core 2 duo cpus. I can't accurately rank core 2 duo as a whole, I can only list where most models USUALLY fit.
It looks like I'm going to have to make some changes to the thread anyways since so many people are complaining.
ExtremeDude2 Wrote:What...wut, I thought that the old phenom II's were faster
That depends on the application and on which specific models you're comparing.
Starscream Wrote:His list is accurate if all the processors involved ran at the same clock speed, generally. A 4.0GHz FX 8120 will beat a stock Core 2 Quad, but not if the Core 2 Quad was running at 4.0GHz.
The list was actually meant to rank them by their average performance at stock speeds, it's just very hard to account for differences between models. Also the FX series cpus were brand new and I had trouble determining where to put them. Since we have so few dolphins users with FX cpus that are willing to conduct actual benchmarks instead of useless youtube videos I had to rely on benchmarks of applications with similar code to dolphin (compilers).
