AMD Phenom II X4 810 & X3 720 Black Edition

[Rob Williams]

Phenom II may have just launched last month, but AMD didn't want to waste time in following-up with their first AM3-based processors. We're taking a look at two, including the X4 810 and X3 720 'Black Edition'. Both offer great performance at their respective price-points, but the X3 became the more appealing chip, thanks to its overclocking ability.

You can read our full article on the AM3-based chips here and discuss it here.

I have to admit, this article didn't go according to plan for the most part, so if I happened to have missed something in the article, please let me know and I'll see what I can do. I'm unsure what it is, but I've been running into the weirdest bouts of bad luck over the course of the past month or two, but I'm confident that's done with. Haha, it has to be, else I might go insane

 

[Doomsday]

Nice review. Now i wish i had waited for AMD Black Edition Processors, atleast they r easier to overclock. i've been reading up some How to Overclock articles n stuff to start overclocking my cpu. i should buy a cpu cooler before i try OCing right?!(Cooler Master V8 is available in stores, wohoo!)

any link which u guys think is the best OC-ing guide for dummies?!

 

[c4xp]

typo

Page 12 - Final Thoughts

As I mentioned in the intro, AMD's main goal right now is to remain price competitive. Their processors are not quite as fast as Intel's offerings
...
consumers always want the best bang for the bang,
...

Should have read bang for the buck no ?

 

[Rob Williams]

Whoops... bang bang! Thanks for letting me know... not sure how that snuck past me.

 

[Moshpit]

Yet again....

I enjoyed the overall article right up to the gaming part and pretty much skipped over it. It's become almost painful to read reviewers CPU gaming tests now considering this bias towards making sweeping statements about gaming in general based on a few first person shooter gaming tests. Is that there is? There's no other gaming genre's? Sorry, that's just silly IMO. The fact is, EVERY REVIEWER who does that is wasting his/her and our time. If you don't already know that FPS games are GPU limited, your in the wrong business, period. Farcry 2 shows a LITTLE CPU dependence, but that's an exception to the general rule that your won't be able to detect any glaring differences if you weight the test in favor of an unrelated component, the GPU by choosing to highlight games that do exactly that.

Games like Company of Heroes, Dawn of War, World in Conflict, Sins of a Solar Empire, Supreme Commander, and other RTS games are VERY popular, and players of these games get NO benefit from blanket declarations of gaming performance based on a genre of games that performs completely differently from it. What you may not know is that these games can benefit FAR more then the games showcased in this and many other reviews. It's almost like there's a concerted effort to conceal the gaming performance of modern CPUs by only showing the games that DON'T benefit from it. Why?

/rant

 

[Rob Williams]

Hi Moshpit:

Thanks for the comments! As I've mentioned in the past, CPUs are not the decider of all things when it comes to gaming. As long as you have a nice Dual-Core chip, you're not going to be held back (unless of course, you want to do heavy multi-tasking, which is where a Quad-Core would improve things a little). As for different genres... of course there are other genres out there, but the first-person shooter seems to be the de facto for those who want to upgrade their computer in the first place. It's the most popular genre among PC gamers, and that's why we focus on them.

From previous experience, I haven't found any game title that has benefited from anything more than a Dual-Core, FPS or not. There are some sim games out there that can use more than two cores, but I've found that they make absolutely no difference in the average FPS, so I'm not sure where the benefit is. In general, I believe that regardless of what genre we choose, we're always going to get similar graphs to these. If you know of games that scale with CPU power, please let me know. I haven't had the chance to test out all of those that you mentioned, but that sim I mentioned above was Supreme Commander (the original).

Thanks again for the comments, I really appreciate them. Next time I have to revamp our testing suite (I basically just revamped this one in late December, so I am not too keen on doing it right now), I'll consider other genres.

 

[Moshpit]

Hi Moshpit:

Thanks for the comments! As I've mentioned in the past, CPUs are not the decider of all things when it comes to gaming. As long as you have a nice Dual-Core chip, you're not going to be held back (unless of course, you want to do heavy multi-tasking, which is where a Quad-Core would improve things a little). As for different genres... of course there are other genres out there, but the first-person shooter seems to be the de facto for those who want to upgrade their computer in the first place. It's the most popular genre among PC gamers, and that's why we focus on them.

From previous experience, I haven't found any game title that has benefited from anything more than a Dual-Core, FPS or not. There are some sim games out there that can use more than two cores, but I've found that they make absolutely no difference in the average FPS, so I'm not sure where the benefit is. In general, I believe that regardless of what genre we choose, we're always going to get similar graphs to these. If you know of games that scale with CPU power, please let me know. I haven't had the chance to test out all of those that you mentioned, but that sim I mentioned above was Supreme Commander (the original).

Thanks again for the comments, I really appreciate them. Next time I have to revamp our testing suite (I basically just revamped this one in late December, so I am not too keen on doing it right now), I'll consider other genres.

I'm forced to strongly disagree with you. If you've tested World in Conflict, you'd begin to see what I'm talking about. In reference to Supreme Commander, if you tested it using the canned benchmark included with it, you learned nothing about the game from that, I can promise you. Simply put, the canned benchmark in SupCom only runs 2 AI's against each other with less then 300 units each and on one of the smallest maps the game has included with it out of the box. A MUCH better test, and at any resolution you care to run it at, is 7 AI's versus the player with max units and an 80km x 80km map. I happen to know from personal experience how ugly that gets on a dual core chip. It's unplayable by the point of 2 hours ingame. On a quad, it remains playable.

Company of Heroes is another one where the canned benchmark was made to satisfy GPU makers and not show real ingame max engine stressing. Download the map called "The Rhine" and play it against 4 AI's on a dual core. It will be unplayable within an hour of starting the game. These points are undeniable. Try these scenarios yourself and you'll see what I'm talking about.

 

[Joachim21]

I can agree with Mosphit! BEnchmarking is quite a things, that can make things look different, than they are in real life.
Maybe, Fottball MAnager 2009, which does many calculation in a short period, and use more cores could be a good test as well!
Beside that, I liked the article, it was correct! Hope to see more test from Rob!

 

[chiefligtning]

nice article

nice review rob but i know you didnt put much time in on the 810 as the 720 but in doing some playing on mine I have reached 3.2 g on my 810 without raising the voltage. and running at that is stable as well as running cool as well