BEST GAMING CPU NOT PHENOM II

This isn't anything new, when P2 came out it's original benchmarks showed that C2Q and I7 are slightly faster in most games when at the same clock speed.

not slightly ....noticably

look at the charts that are frame by frame

http://www.hardocp.com/image.html?image=MTI0MTI0MjEwMUh...

y didnt they put the 955 in there, y didnt they put the 945, the 905e

funny how the i7 920 at the same clock speed as the x3 720 in "crysis" the holy grail of performance only averaged 3.5 fps more than a cpu that that has 1 less core, 5 less threads and only 95w tdp compared to its 130w oh and a $280 cpu scoring only 3.5 fps over a cpu that costs $119, you intel fanboys really like giving those millionaires and billionaires your money, oh and did i see wrong or did they get the same 6fps min, is $160 worth 3.5 fps you tell me.

y are intel fanboys always so ansy for a fight , ive never seen a thread saying
(is $160 worth 3.5 fps you tell me?)

Intel Core i7 920, Intel Core 2 QX9650, AMD Phenom II X4 810, & AMD Phenom II X3 720 BE

So the i7 and the best core2 beats the mid range phenom X4 and X3 costing half or less. Even then, in NO benchmark does the intel's allow for higher settings.

WOW GOOD JOB INTEL.

Like whats already been said, theres very little difference between todays chips, but some people are stuck on yesterday, and P1' perf, which admittedly wasnt anywhere near Intels in gaming.
Thats changed, and tho they still havnt topped Intel, theyre right there next to em.

What I find funny is, alot of people have never forgiven AMD for the 939 skt thing, and the subsequent let down of 65nm and P1. Id just like to point out, I think its time for all those former Intel fanboys to forgive Intel for P4 heheh

Your missing the point. The value of this article are the REAL-TIME line charts.

FRAME BY FRAME BY FRAME

http://www.hardocp.com/image.html?image=MTI0MTI0MjEwMUh...

PII is on the bottom

Both Intel solutions reach close to 3Ghz. Both cost much more individually, as well per platform (cpu included). Why not use the money saved, if it makes no difference, and use a better gfx card, like most people would do?

http://www.hardocp.com/image.html?image=MTI0MTI0MjEwMUh...

Take the only tri-core out of that (the green line) and what does it look like? Yeah it's a lot closer now isn't it? Also, note the 9650 scoring the lowest fps on that particular chart.

It doesn't matter that they are all clocked to the same clock speed when the AMD's are crippled in other aspects. The 720 lacks a core, the 810 has crippled cache (very important for gaming). And they put 2 of the best intels up for comparison?

What point am I missing here? That faster and more expensive cpu's perform better? I don't think I'm missing that point, what point are you missing?

Sorry, didnt know about the clock speeds, I avoid [H] as much as I can.

The thing here is, having them all at the same clock speed means absolutely nothing.

The x3 lacks a core and that clearly shows in the majority of benchmarks. It also quite clearly shows how quad cores are better for gaming regardless of what you see in benchmarks. I've had plenty of each and I know (and I'm sure anyone who has upgraded from dual to quad knows), quad cores are far better in gaming that dual cores.

I just don't get why they would put the 920 and q9650 up against the 810 and 720. What was that supposed to prove except what we already knew which is more cores and more cache helps a lot in gaming. More clock speed helps a lot too, but at least try and make them consistent while doing a review over various parts.

In an earlier thread, someone asked to rate the importance of clock vs FSB vs cache. Its hard to pinpoint exactly, as some games apps nneed the cache, while others dont, tho gaing, its a good idea, and I pointed out in that thread, you shouldnt worry about cache on a full cpu, but the lessor lil brothers is when youll see its lack show in perf, not in the full cpus. I thought everyone knew that?

And look at the average, a mere 2-3fps average isn't noticable. I also see a number of spots where the Phenom is on top in that frame by frame.

Try this chart http://www.hardocp.com/image.html?image=MTI0MTI0MjEwMUh...
Intel Core 2 Quad Q9550 Yorkfield 2.83GHz LGA 775 95W Quad-Core Processor Model BX80569Q9550 - Retail $220 FS Easy OC
http://www.newegg.com/Product/Product.aspx?Item=N82E168...
AMD Phenom II X4 945 Deneb 3.0GHz Socket AM3 125W Quad-Core Processor Model HDX945FBGIBOX - Retail $225 FS
http://www.newegg.com/Product/Product.aspx?Item=N82E168...
So we know all the CPUs will all make at least 40FPS. Then we have to ask how often they make at least 40FPS. The line chart, without AA which taxes CPU most, shows the PII to lag most of the time.

We realize that C2Ds are slightly better, not alot, but look, at least compare the right cpus together, thats the point were trying to make. Use a P2 955 vs this cpu, and itll be close, very close, just like the prices are as well.
[H] comes up with some off the wall crap at times, and even their regular stuff isnt consistant to their own methods, which just opens the doors to their opinions, as opposed to real facts. I dont trust them enough, due to their inconsistancy, or comparing a lower level cpu to a higher level one like in the link youve provided.

I cant remember all their failures, even with their own testing methods, others here Im sure do tho. Go to the gpu section, and ask. Therell be a purple apeish thing that can answer alot of that heheh

Link? You must be color blind. http://www.hardocp.com/image.html?image=MTI0MTI0MjEwMUh...

Probably the only thing I like about [H] reviews are those graphs. I've said it before: just because the minimum is really low means nothing if it was only there for 2 frames. You need to see the framerate changes across the entire run to get the full picture. [H] have obviously never heard of demo recording otherwise they would have had more consistent L4D results. What they have there is meaningless.

And the graphs are the only facts, but you also have to be able to read them as well. I like their ideas, but theyve distorted, and politcized them so bad, to me it doesnt matter anymore what they say or find

Yea they need to make the full-sized graphs wider so we can see what's going on more easily. When the framerate jumps up and down all over the place they end up being a picasso.

Amen to that, all I see is a bunch of lines running over each other. Its impossible to glean any useful info from those graphs.

Makes it easier to talk up or down what you want to doesnt it? I remember someone pointing out how in their graphs, an ATI card was bottoming out, and an nVidia fanboi was proclaiming how bad ATI was, and just look at those graphs.
Ape came along and explained to said fanboy, theres intervals in between, where it shuts down altogether, thus the dips. Woulda been nice of them to explain that, but someone had to ask in their forums, (and was probably kindly expelled heheh) before they admitted it

FPS (frames per second)

681+ seconds on the chart

use the zoom feature of your browser

seems simple to me C2Q White UP, PIIX4 Red DOWN

Maybe , since its really a lowly i3, they shouldve used a celeron to compare? Whats in a name?

Zoom feature doesn't do anything but make the lines bigger, they still cross over exactly the same, only you can't see the rest of the graph. The graph needs to be redrawn wider, that is the only solution. Height probably won't matter. But [H] won't want to make their dubious conclusions more obvious so nothing is going to change.

Yea real tough call ..

See those drops? Thats where it stops and starts again. Funny how only the P2 shows up so low? Look at the intervals on the drops, theyre the same amount of time. A little doctoring perhaps?

.

According to your logic, you can also get a cheap Pentium Dual Core.

It only has 1 less core, 1 less thread, only 65w tdp compared to 95w oh and a $119 cpu scoring only a couple fps over a cpu that costs $60 (half its price).


It's called paying a premium for extra performance. That's why AMD has Phenom II x4s and Intel has higher end Core2Quads and i7s. Nobody actually needs them in terms of real world performance in gaming, but they're still good to have.

This has to be one of the best misleading posts ever done. He takes the time to link and mention the specs of the 945. I then think the graph must be a different one, one that includes the better 9 series PhII. Nope. Still got the x3 and the lowly 810. I'm with jenny on this one, this thread and that article don't really prove anything. Lower end three and four core CPUs are slower then higher end quad and 8 thread CPUs. Wow, hell of a news flash there folks. I think my world is coming apart now.

For those of us who don't have money coming out our @$$e$, the fact that the x3 and x4 in any flavor comes close to the best of the C2Qs is great news. This allows us to get a "good" CPU, and spend the money saved on a better GPU. If you only have a $400-$500 budget, you can't afford to get a $220 CPU. Buying an X3 and putting up with slightly slower results IS great news.

This is the problem with a bunched-up graph. The lines overlap and only one can be on top. Whether intentional or not, the graph is misleading because you can't see the other lines behind and anyone who doesn't stop for a moment to think about it won't think that maybe the QX and i7 are in behind there as well. They are based on minimum framerates shown below the graph.

What do you expect someone who cites [H] to begin with?

Like I said, its but another example to avoid [H]. I bet if one really checked, the particular HW that [H] wants to crap on is always the visible one

^+1
Really shocked to see this thread and that article that it links to...

X3 720 - $120
X4 810 - $160
Q9650 - $320
Core i7 - $280

How can those people even write an article comparing CPUs that dont even cost half the price of one with other  

A serious let down...

Well, I essentially got my i7 920 for just under $200 AU.   government rebates ($392 upfront, will be getting the money back in my tax return)

They shoulda included the EE models as well heheh

Id love to get Kyle on tape, gaming on 2 different rigs, 1 P2, the other C2Q, and then, have him give his expert advise, but without him knowing which was which

I'd also like to see them also compared to the $103 3.1 Ghz Phenom II 550 with the 2 other cores enabled

I can move this into the sticky if you want to keep this all in one place so that it doesn't get buried overnight by help threads. I just hope that the merging function works nicely for once...

As stated, I also avoid [H]. Their editors cannot take any criticism and will down right be rude to anybody they want to on their message board.

Toms, Guru3D, Benchmarkreviews, Anandtech all take criticism openly and objectively and these 4 sites I've seen, take forum suggestions and implement them on their testing.

Hell, Toms asks!! I may not agree with them, or their findings all the time, or even their methods on pccasion, but Toms somehow always delivers

I don`t live in Australia > ... still, really nice buy... seemed like me when I purchased 2 OC 4850 for 190$ cad 165$ US > 

Rebate Hunter we are > 

IMO, it`s all that matter.

If someone come for a gaming rig, I don`t put the most expensive stuff together, I purchase the best ratio price/performance...

Right now it`s the X3.

Well, their latest article claiming comparing CPU with different graphic cards is a bit... how to say that politely... useless...? :S

Comparing ATI/AMD with NVIDIA/Intel... lol, way to go for a SERIOUS and objective compare...  

Which article is that? I haven't read many of the latest ones. Some of the reviewers here are definitely better than others. I don't think they intentionally skew the results but there's often some discrepancies. They certainly don't get paid for any of the reviews, that's for certain (as in BoM don't, the reviewers obviously get paid by BoM).

To be fair, I have seen good [H] articles, especially their PSU ones. I don't always agree with the conclusion they come to, but at least the data is there. They do seem to always put down ATI/AMD, but I'm not 100% sure if its because the products are/were that bad or they are biased against those companies. I do think its important to test the way they do, it provides another tool we can use to compare. The problem is we need to make sure we aren't a tool and swallow anything they feed us.

The important thing to remember is that regardless of whether a site is biased or not none of them are paid to write advertisements. They are certainly pushed to adjust their analyses but never paid.

Oh, and if this thread seems to disappear then it's in the discussion sticky. It's omre about [H] than AMD/Intel now anyway.

The PII X3 720 has 6MB L3 Cache so it is relevant. Right now the PII X3 710 and the

PII X2 550 BE are same price $100 FS @ newegg so I don't see the point of that.

Microcenter currently has the E8200 retail for $106 shipped. Having four cores

doesn't mean anything if your min FPS crap out 40% of the time.

Ok Penryndo, let's try again.

http://www.hardocp.com/image.html?image=MTI0MTI0MjEwMUh...

Take out the green, it's a tri-core.

From 1-61 seconds, the 810 is a clear 10-15 fps faster than the i7. You've just played a whole minute at either 30fps or 20fps, what would you prefer?

Now look at the end, and you'll see once again the 810 has a clear lead. The last 3-4 minutes has the 810 ahead almost without fail.

In the middle, the i7 starts to get some high numbers, especially from 241-461 seconds. What actually is happening here is, those high middle numbers are hugely changing the overall fps in favour of the i7.

Over that 11 minutes however, the 810 has a consistently higher fps total for much of it. What is the explanation for that?

So you're kinda right Penryndo - average framerate doesn't tell the whole story. Because, in this case the i7 comes 0.9 fps higher in total, even though if you were playing it on both the 810 would prove to be a better experience for longer periods of time.

The 810 is better at the start and end, and the i7 is 'more' better in the middle where the fps isn't such an issue. We've seen this before with the i7 - it's why it owns at extreme low res gaming. This graph is good proof of the i7 failing quite hard at extreme high res gaming however, because there is no other way to justify it losing to an 810 over such a long period of time.

They are running at different settings though. Gamer for the Phenoms and Enthusiast for the i7 and QX. This is where [H]'s "highest playable settings" rubbish comes into effect.

X3 720 will probably be the fastest if your trying to stick to a budget because you'll be able to buy a more powerful graphics card(since your spending less on the cpu) which will make a bigger diffference to the framerate you get.