Cyberpower’s Gamer Dragon: Can AMD Bring The Game?

[Guest]

Here is a good article on how to overclock a Phenom:
http://www.microdots.org/ramblings/node/30

cyberpower's overclock is a joke.

 

[da bahstid]

Maybe I should have specified more obviously, but I actually do more or less trust these numbers, with the understanding that the selected benches tend to cater to i7's strengths. The games selected are all known to wreak havoc with cpus. Far Cry 2 happens to be specifically threaded to take advantage of i7 hyperthreading, for example...that's a win that isn't any surprise. Then go out of your way to run heavy X-fire/SLI systems at 1024x1280 with AA turned off to remove graphics from the equation as much as possible. That Phenom II more or less broke even in the last couple games with this test setup is actually more the surprise to me.

Do another setup with graphics-oriented games that aren't as cpu-bound (which is most of them, really) and run them at the more practical resolutions for such graphics card setups (meaning 1050x1680, 1200x1920, and 1600x2560) and the gaming results would look much more in line with benchmarks we've seen everywhere else. Here is where one can say the test bias is. Doesn't change the fact that Phenom II ultimately only breaks even with Core2 while Nehalem is the next step forward. The differences seen here today may not be common across today's games and in most people's setups, but they probably do a good job of representing the degree to which i7 will eventually pull ahead of both Phenom II and Core2. This is still massive progress for AMD though, given that prior to Phenom II they were a tough sell even as budget components. Right now they're the budget kings. Intel has no answer for AMD's $120 720BE triple-core. It would have been crazy talk to suggest such a thing 2 years ago.

 

[Seikent]

farcry, crysis and world in conflict are clearly cpu limited, there are a lot of other games that could make the 4890s shine.

 

[noobinberg]

We already know that the i7 beats pII...that is clear...I liked this article because it showed me just how much better Intel is than AMD right now...now I know that I could build that pII system for far less money than the i7 one...but why would I do that? Everytime I build a system for myself I use the best parts that my budget can afford....back in the 90's it was an AMD cpu...but as soon as core2 came out...i switched...if you wanna save money and sacrifice performance get a phenomII...if u want faster and can afford it get i7...and furthermore, if they oc'd that i7 like the pII it would have been a major blowout!!!!!!!!!!!

 

[playerone]

From just a fresh culling other benchmarks does it just seem weird that the author didn't even wounder about the data?, Or wouldn't think the readers would wounder and maybe do his utmost to unravel the mystery or explain? If this site is just going to head into "ZiffDavis" world then it starts to put so much more in question. How relavent is anything else that we look to fro accurate and informative information. Is the Author not going to look into this?

 

[playerone]

Note to self: Dont post in such a hurry! Or at least spell check LOL!

 

[Mithrandir95]

Wow, can't believe how biased this article is! Why benchmark games on top of the line machines with just about every test not using AF, and at the best 4X AA? And NVidia biased games at that. How about benchmarking it like how someone who buys this system is actually going to play it, at high res'es with high 8 or 16x AF, and 4X 'or better' anti-aliasing. Maybe then we'll see a system benchmark and not a CPU benchmark. We get it already, the Core i7 is faster than the Phenom II, but it's also generally more expensive, lets actually see what happens if that extra cash is used to upgrade other areas of the system.

 

[dmv915]

[citation][nom]armistitiu[/nom]It's articles like this one that make me lose my trust in Tom's Hardware.Comparing home build vs retail PC from Cyberpower is outrageous. One thing: is a big difference in price. Second: Why would you use 790X which has pcie 8x for Crossfire?I'm not saying Phenom2 is better than i7. i7 beats phenom2 955 in most of the tests. Ok. But how in the world can you compare gaming performance on two systems that are different in so many ways.Allright. I'm out of here.Call me fanboy all you want but you've let me down Tom's. Now even Anandtech's tests seem to be fairer.[/citation]

Running cards in x8 vs x16 has negligible difference on the performance of the system, only when we get down to x4 can you really see the difference.

 

[dmv915]

Disregard my previous comment. I just realized I was pulling information from i7 instances. Test done with a p45 and x48 really show the difference between running crossfire on x8 and x16 and that's with pretty low end 4850's.

 

[scooterlibby]

Maybe you don't have the time, but seriously, do some regression analysis if you could.

 

[noob2222]

Here is my take on this article. USE PROPER STATEMENTS.

First, the title. Cyberpower, Can they represent AMD? FAIL

This article has to be one of the worst written articles I have seen in a while, but seems to be more common lately, esp after the SBM Intel build contest.

This statement alone makes me want to go hug my toilet.
Page 3, test and system setup

So the reasonable overclock of the Cyberpower Gamer Dragon at 3.6 GHz will do battle with the reasonable overclock of the SBM machine at 3.44 GHz. Once again, this should make for a great battle, and represents a best-case scenario for the Phenom II-based system.

BEST CASE SCENERIO MY ARSE. Sub standard 790X mb (probably one of the slowest on the market), sub par memory with crappy timings (spend $10 on better memory, check the memory article on the x4 and see a difference of 6-10fps), hdd picked for its size (twice the cost of the SBM) but you try to shove the idea down everyone's throat that this is the best AMD can do, with this Cyberpower article.

The results are as plain as day: the Phenom II can bottleneck a theoretically-superior graphics system in modern game titles, and even a stock Core i7 will allow those games to stretch their legs.

This article should be labeled after Cyberpower, leave most of the AMD biased statements to an AMD article, not this load of crap pretending this is the best system AMD can muster.

My conclusion, feed a decent cpu some crap, you end up with crap. Feed a better I7 some crap, you get a little better crap.

 

[reasonablevoice]

I don't know how your Phenom build cost you anything near your Core i7 build. Part of my job entails configuring, pricing, and assembling new computers for customers, the phenom platform should be much cheaper.

 

[reasonablevoice]

Also, you've been very selective with the games you've chosen and not applied AA/AF in most of them. I think you'll find things much different the higher the resolution and the higher the filters. Most enthusiasts are gpu limited, where an i7 will not provide much benefit.

 

[ARCHER0915]

This was not necessary:

What did we learn from the Cyberpower Gamer Dragon? A heck of a lot, and the points were more dramatic than we expected.

First off, we learned that new game titles will definitely see a benefit from the high CPU power that the Core i7-920 brings to the table. The results speak for themselves, so we can’t really be accused of brand favoritism for choosing a Core i7 system over a Phenom II in the System Builder Marathon series. The results are as plain as day: the Phenom II can bottleneck a theoretically-superior graphics system in modern game titles, and even a stock Core i7 will allow those games to stretch their legs.

Are we saying that the Phenom II is a bad buy? No. But Phenom IIs shine most brightly where they make sense: in low-cost systems that a Core i7 can’t reach due to comparatively-high platform costs. If we were going to put together a Phenom II system, we’d use a lower-cost AM2+ motherboard, DDR2 RAM, and a couple of Radeon HD 4850 cards to better realize the configuration's cost advantages. That would be a formidable low-cost system, and would stand well against any Core 2-based offering from Intel
.

But back to our contender, the Cyberpower Gamer Dragon:

Side notes like that gives Toms a bad rep in some circles. But back to our contender???? what kind of crap was that???? Other than that the review was pretty good and the Cyberpower is overpriced as you so duly noted.

 

[Guest]

It's amazing how much FUD is spread on tech blogs between AMD and Intel fanboys. When AMD wins, Intel supporters claim bias. When Intel wins, AMD supporters claim bias (and the same for ATI vs. Nvidia). Well, I'm sorry people, but right now since the i7 came out, Intel rules the roost, and it's really been that way since the Core 2 series came out (quad or duo - and yes, much of that "win" was because Intel chips overclocked much higher than AMDs). So when AMD gets back into the swing of things (hopefully) to knock Intel around like the good old days, you can stop ya whining. Until then, suck it up like a man instead of a whineyassed little b!tch.

 

[Guest]

@Fudrucker: Nobody contests that a Core i7 would be as good (and sometimes better) than a Phenom II in pretty much everything, what's being disputed is the outrageous margins that this carefully crafted article is suggesting it would be. They clearly went to quite a bit of effort to come up with this horse-sh.it, as randomness wouldn't produce these kind of skewed margins, clearly some thought and planning went into this, with the intent of producing this outcome.

 

[playerone]

Please remove the "cyberpower" review. Very inaccurate crap here.

 

[tychoblu]

Quote from SpadeM (and most other AMD fan boys) on the 1300$ build remarks:
"I wonder how exactly does the selection of components go. I mean it seems that there's some attention given to the forums to be politically correct, but that's kind of it. For $1300 a Phenom 2 + micro AM3 board + 2x4890 in crossfire is a much better solution so .. why not choose the better option?"

This article addressed all those crying AMD. And yet they still cry. QQ

Thank you tomshardware.

 

[eddieroolz]

I still think that the game selections could have been better.

I get bombarded with a lot of "Runs great on CORE i7" ads when playing CoD:WaW. It's a no brainer that a E7200 can outperform a Phenom 9950 in many games, just because game developers heavily side towards Intel.

 

[goose man]

For everybody who have doubts that ATI 4890 is stronger than GTX 260 Core 216 (especially mcvf) this article in Anandtech will confirm it.

Basically that article is comparing 4890 with GTX 275, but you can see in the chart the GTX 260 core 216 is included, and it is consistently weaker than 4890.
Sometimes 4890 even stronger than GTX 275, but they (4890 and GTX 275) just trade blows.

 

[indigoccs]

AMD sucks, stop crying you cheap bastards, buy intel until AMD gets on their feet, but now their in a hole with a year old performance cpu...

 

[indigoccs]

"I wonder how exactly does the selection of components go. I mean it seems that there's some attention given to the forums to be politically correct, but that's kind of it. For $1300 a Phenom 2 + micro AM3 board + 2x4890 in crossfire is a much better solution so .. why not choose the better option?"

WOW where is he now???

 

[Gruntz]

I did my own testing now @ Far Cry 2 and your numbers are completely wrong. I do not know how you did this but telling readers this is the "best case scenario" for AMD ist just plain bullshit.

955BE @ 3.2ghz 2400NB 6-6-6-22-26 1T @1333 ONE 4870 1GB windows 7 64 caviar black 1TB

Far Cry 2 medium ranch DX10(it does not specify in your article but you used it in your old 955BE article: http://www.tomshardware.com/reviews/phenom-x4-955,2278-9.html)

very high 1920X1200 70.5fps!!! with only ONE 4870 1GB, you only managed 62, had 2 4890s AND OC to 3.6ghz while I tested with only 3.2ghz.

This setting was still gpu limited with ony my rig, with 4890 crossfire it should a lot faster.

To simulate a fraction of the performance advantage of 4890 crossfire, I lowered the resolution to 1024X768

FC2 DX10 1024X768 very high ranch medium

-> 89.9 fps

So I have no idea how you can come up with those numbers. 1920X1200 is clearly gpu limited on my system with a 4870 and even though my cpu is not overclocked, I am getting higher fps than you guys with two 4890s.

 

[ewq108]

I think there're something wrong with the ATI Driver , 4890 in CF cannot be this weak >.< .

 

[kevin1212]

And the annoying part is that they are well aware that something is wrong, but still decide to use the numbers just to prove themselves right. How can you call yourself professional when you do something like that. Toms Hardware was easily the most informative and knowledgeable site when it came to computer hardware, but they are really beginning to disappoint me. The only real reason I see here is for this poor info that they are finally on Intel's payroll.