Socket AM3: AMD's Phenom II Gets DDR3

[JumpKickJoe]

[citation][nom]raden_muaz[/nom]I've been waiting for so long for this phenom ii to come.At last, phenom beats i7 in some kind of way.Anyway, I never bought Intel for years because:http://stopthewall.org/worldwideactivism/982.shtmlhttp://www.inminds.co.uk/boycott-intel.html[/citation]
Now that is news to me! I'm a AMD fan all the way, seeing the improvements the Phenom II have are making me weak!!! OOOOOh I can't wiat for the AM3 platforms to come out!

 

[squatchman]

The game benchmarks are odd, but it isn't like the 920 didn't win just about every other benchmark. Add to that the fact that they ran the memory in two channel mode instead of three?

If you see the 3dmark GPU scores you will note that the i7 is also low, where all the CPUs should have had equal scores. Could there have been something wrong with the setup? What kind of power supply was used?

 

[roofus]

[citation][nom]jameskangster[/nom]Ok, I must be missing something here. I compared the AM3, AM2+ reviews from Tom's Hardware and Anandtech (http://www.anandtech.com/cpuchipse [...] spx?i=3512), and the results seem to contradict each other specifically relating to the game benchmarks. I understand there are subtle differences in the configuration setup, but i7 performs not so favorably according to Tom's Hardware's benchmarks, whereas at Anandtech's review, it pretty much stays on top in every benchmark.[/citation]

I was thinking the same thing. I think those who have followed along know the AM2 P2 is pretty well aligned with the mid range Yorkfields and it looks like the AM3 will be trading punches with the likes to the q9550 and q9650 Yorkfields. Outside of that, its common sense the AM3 is going to win benchmarks that the Conroe could beat i7 in. I really wish they would have dropped in a Q9550 or Q9650 for comparison on the charts.

 

[sohei]

man in games for now maybe.. and maybe forever this are the performance ...anyway i7 is a good cpu in most cases great

 

[ramper1]

Its possible those AmandTech benchmarks weren't even using DDR3 like some posters mentioned. The article if misleading and reaks of fanboism, "Dont buy DDR3 unless you are going i7" "Thanks Intel for not ripping us off anymore" speculations and 'predictions' at every corner the guy who wrote it is a douche.

 

[ramper1]

[citation][nom]ramper1[/nom]Its possible those AmandTech benchmarks weren't even using DDR3 like some posters mentioned. The article is misleading and reaks of fanboism, "Dont buy DDR3 unless you are going i7" "Thanks Intel for not ripping us off anymore" speculations and 'predictions' at every corner the guy who wrote it is a douche.[/citation]

 

[blackened144]

[citation][nom]raden_muaz[/nom]@t85usI do expect Intel to excel far better than AMD (I'd love to buy an i7, but I just can't)About the massacred nation - I'm a Malaysian - they massacred my muslims brothers in Gaza.[/citation]
Your Muslim brothers may have been massacred, but they had it coming and if you ask me, they got off WAY too easy. Now t85us and raden_muaz, keep your political commentary to yourself, most people think it is perfectly acceptable to retaliate against people launching rockets into their country.

 

[neiroatopelcc]

So ramper you're saying, twice, that choosing ddr3 when you could've chosen ddr2 for a fraction of the cost is a good choice?

I've always had the motto that quantity is better than speed when it comes to memory. I'd rather have 8gb ddr2-1066 than 3gb ddr3-1333. But perhaps that's just me...

 

[ramper1]

Should get some edit/delete options on these boards heh, the amandtech article comes across as trying to sway the reader to buy Intel and gives barely any hard data. AM3/DDR3 is still in its infancy so how can anyone state that it does not need a feature. The best reviews and conclusions imo will come from the hard core enthusiasts who manhandle these components all day. But yea neiroatopelcc you said it yourself its what you prefer.

 

[neiroatopelcc]

[citation][nom]ramper1[/nom]Should get some edit/delete options on these boards heh, the amandtech article comes across as trying to sway the reader to buy Intel and gives barely any hard data. AM3/DDR3 is still in its infancy so how can anyone state that it does not need a feature. The best reviews and conclusions imo will come from the hard core enthusiasts who manhandle these components all day. But yea neiroatopelcc you said it yourself its what you prefer.[/citation]
Somehow for me it's a tad hard to conclude what you're trying to say. Are you saying that it is better to buy a bit of ddr3 memory instead of more, or cheaper, ddr2 because the forums need an edit button? or because am3 is new? I'm sorry, but it's really a nonsense message you left there!

 

[ramper1]

http://www.xtremesystems.org/forums/showthread.php?t=217281

 

[ramper1]

Its up to you man... anyway you can keep up to tabs with some ddr3 and am3 here http://www.xtremesystems.org/forums/showthread.php?t=217281

 

[neiroatopelcc]

[citation][nom]ramper1[/nom]Its up to you man... anyway you can keep up to tabs with some ddr3 and am3 here http://www.xtremesystems.org/forum [...] p?t=217281[/citation]
You really don't like answering questions .... but it's okay I suppose. You're probably just a politician or something.

 

[Pei-chen]

What is wrong with your i7 setup?

"Intel's Core i7 920 needs to be overclocked in order to serve its purpose in a gaming machine. Here we see it again slowing things down versus AMD's lineup"

So a tri-core X3 8750 @ 2.4Ghz from a year ago managed to beat i7 quad core @ 2.66Ghz?

 

[cangelini]

[citation][nom]Pei-chen[/nom]What is wrong with your i7 setup?"Intel's Core i7 920 needs to be overclocked in order to serve its purpose in a gaming machine. Here we see it again slowing things down versus AMD's lineup"So a tri-core X3 8750 @ 2.4Ghz from a year ago managed to beat i7 quad core @ 2.66Ghz?[/citation]

After building the system up, breaking it down, building it up again and re-running all of the i7 920 tests, the gaming scores came out the same. We questioned Nvidia on its latest drivers, but everything looked normal to them. It didn't seem to be a configuration problem--the i7 still dominated in the productivity-oriented benchmarks.

 

[blibba]

Definitley an issue with those i7 scores. There's no way it's worse than an e8 series.

 

[sohei]

people come on!!! intel it.s the god of cpus ? is just a company where work people and people does not do always the perfect products ...and please you want just one company ? one company means communism

 

[ARCHER0915]

I for one am disappointed in the review as it does not address the one question on the minds of a lot of those going through this cycle with AMD:

I want comparative real world benches done with DDR2 to DDR3 are there advantages-disadvantages same cpu matching boards as close as possible and then and only then does this become a real comparison. We already know you can pump the data in and out but while the data is in memory it must be manipulated and that is where the difference is.

 

[adaml007]

Someone already mentioned it once but you really need to use the triple channel on the i7. Using 4 GB (2 GB x 2) Corsair Dominator DDR3-1600 8-8-8-24 @ 1.65 V on the benchmarks is really not fair to the Intel Rig.

 

[cangelini]

[citation][nom]adaml007[/nom]Someone already mentioned it once but you really need to use the triple channel on the i7. Using 4 GB (2 GB x 2) Corsair Dominator DDR3-1600 8-8-8-24 @ 1.65 V on the benchmarks is really not fair to the Intel Rig.[/citation]

Updated the test config to reflect that the Intel platform indeed was using three of the Corsair modules--the AMD, naturally, only used two. Thanks Adam.

 

[hoodsax]

"So, using a mechanical pencil, we bent the two offending pins back and forth until they snapped off. The chip now fit into our AM3 test platform, albeit not flush due to the metal nubs where each pin broke."

This would have been amazing if it had worked and might have redeemed THG (a little) if it had. Although, it didn't, so it is we-tall-did that you did this and you shouldn't have even mentioned it.

 

[hoodsax]

Maybe cut off all the pins and try to put it in an X58 motherboard.

 

[kknd1967]

That 3DMark GPU score made me laugh. Please, THG, either this article's editor or 3DMark programmers should be fired, for giving such different GPU scores on the same video card, even if this is across different platform.

 

[kknd1967]

I really have to post again after reading all the gaming benchmarks. Dear Chris, unless you are doing internship at THG, please learn to do a reasonable benchmark. Particularly look at gaming performance compared to other sites such as Anandtech. If you could not justify your numbers, you have to rethink what you did wrong.

 

[hannibal]

Those differencies can allso tell that on some systems thins prove to be different. It can be an bios issue, or anything, but for some strange reason, some hardware parts does not fit so well together. Any PC builder can tell that for some unknown reason some combinations just don't work so well...
So IMHO the testing can actually be right, but it's good to see different combinations, so you can expect that in most cases i7 should be faster

But all in all the most interesting part is how the dd3 affect the Phenom II compared to dd2 variant. So far it seems to be a tiny little bit faster and allso more energy efficient. Seems to be in scale, in what was predicted after Phenom II was released.