Phenom FX in the Works, AMD to take Another shot at...Kentsfield

[fullinfusion]

Depends on what you are doing realy? Some apps might need alot of bandwith or memory etc, some dont, all depends.

Idk about that, AMD architecture has always been infront of intels, even intel its self admitted this, why you think they are goin the same architecture to AMD's? I think once the Cache sizes are equal to a intel at this time, not including the new 32nm CPU's that intel will bring out, that you will see that the performance will be as good if not better, and yea maybe 4 cycles included, not sure tho.

All we both can say is, it might or might not, in the end we will just have to wait, all im goin on is past tests and how it used to be.

Thanks Melvis.... i appreciate that man... AMD all the way.... sorry i just hate to see them loose in the cpu department.... god only knows they pulled there sh^t together with there new GPU's against Nvidia

 

[blueskynis]

You guys filled my mail!

 

[Tatty_Two]

Mmmmm i like this war goin on between these two^ hehe i like to see the end results , but i have to admit that the Wile E machine should have a advantage because of the amount of Cache the CPU has, 3 times as much right?

Thats the only reason intels are going so well, the amount of Cache they have on there chips, and its very expensive to do this process, but for intel its not, AMD it is.

Bit unfair?

No thats not the entire story by any means, if you put that much L2 cache on an AMD it would make little or no difference, they are completely different architectures that require completely different parameters, AMD's architecture does not need the extra L2 where as Intels is enhanced by it, I suppose my point is, the extra L2 in itself does not make the gains clock for clock between the 2 architectures, it;s the efficiency within those architectures that determine the speed, for example, AMD is/always has been very strong on floating point calculations and memory intensive processes (AKA on Die mem controller), Intel is much stronger on arithmatic blah blah.......SceinceMark 2 sees the two of them more closely matched clock for clock as that bench tests memory and various CPU calculations.

Sorry long winded answer to a very small point!

One last point, you say, AMD's architecture is more advanced, that could be true, however, it means nothing if it cant compete, and sadly, across the board it cant compete.....now before you think it (if you havent already!) I was brought up on AMD, I love AMD and I will always go back to them.....when they become more competative.

 

[Wile E]

na like i was saying whil running msn in video and audio a friend running 3.6 to my 3.01 was so close with the both of us running the same program at the same time.... he even said( you @ 3ghz to my 3.6 gives the intel a good run for it's money)..... im off to bed guy's.... i'll come back on with weird mem timings and fsb to up the points a bit.... lol.... it's all good.... but Wile...... I do have a 750sb mobo on the way..... the bottle neck is now looser.... lets play when it comes... ok.... lets just see what becomes of the Phenom then.... may not be as good but i bet ya it's real tight! night guys

Yeah, no problem. I love running benchmarks. it's fun to see how different setups compare.

 

[Melvis]

well only time after the movie is done burning.... after that i ungang the memory and she smokes rite along utilizing all four cores converting.... so all in all around the time i pop the dvd in to burn till the end to watch on the xbox 360 through the media extender it's 45min

Thats still half the time of what mine can do, and its good to know it uses all 4 cores, not like half the games out still that only use one still =/ (Supreme Commander) Silly design in that game.

good stuff man

 

[fullinfusion]

ganging the memory while concerting a dvd file takes around(dont hald me to this) but around 12min's longer compared to unganing it... that's the one thing i love abought this 9850 @ 3.15GHz 24/7.... as now the new bios came out..... heck 3.05 was hard to keep stable on the 0801 bios.... Asus made a huge improvement on there new bios to let me clock that for a 24/7 system.... i however am still testing at higher multi with lower fsb's and 1066mhz setting mode for memory.... tests are showing better performance day by day...

 

[Wile E]

You guys filled my mail!

Go into your User CP for this site, go to Edit Options, scroll down to Default Thread Subscription mode, and change it to not email you when somebody posts in one of your subscribed threads.

 

[fullinfusion]

Thats still half the time of what mine can do, and its good to know it uses all 4 cores, not like half the games out still that only use one still =/ (Supreme Commander) Silly design in that game.

good stuff man

yeah i hear ya man.... i down core in GOW to 2 cores just to save the power that im otherwise wasting while playing..... i had a 6400 clocked out to 3.5GHz and she didn't even come close to converting movies like this Phenom does....ahh let there be light at the end of the tunnel for the 45nm fx processors.....

 

[Melvis]

Thanks Melvis.... i appreciate that man... AMD all the way.... sorry i just hate to see them loose in the cpu department.... god only knows they pulled there sh^t together with there new GPU's against Nvidia

Know Problem dude, ill go AMD for a long time, but lets say in 5yrs time they still havent got ontop of things, i might go intel =/

But time will tell, they have done it b4

 

[fullinfusion]

Go into your User CP for this site, go to Edit Options, scroll down to Default Thread Subscription mode, and change it to not email you when somebody posts in one of your subscribed threads.

Wile your pix is like a bloody eye magnet every time i see it!!!! it's like staring at me every time i see a post from ya lol....

 

[fullinfusion]

Know Problem dude, ill go AMD for a long time, but lets say in 5yrs time they still havent got ontop of things, i might go intel =/

But time will tell, they have done it b4

umm 5yrs? i doubt it.... give it mabey 8 months lol.... but who know's.... im AMD all the way... cheers mate.... off to bed...-out-

 

[Wile E]

yeah i hear ya man.... i down core in GOW to 2 cores just to save the power that im otherwise wasting while playing..... i had a 6400 clocked out to 3.5GHz and she didn't even come close to converting movies like this Phenom does....ahh let there be light at the end of the tunnel for the 45nm fx processors.....

I really hope AMD does pull a rabbit out of their hat with Deneb. But overall, I think we may have to wait for K11 (or whatever the next architecture is called) for there to be a significant shift in performance.

Let it be known, I am not an Intel fanboy. I will switch back to AMD the instant they take the OCed performance lead. (Actually, I still have an AMD rig with a 6400+ in it. I just wish the damn board supported Phenom. )

 

[blueskynis]

Fudzilla said Deneb will be K11, I think?

 

[Melvis]

No thats not the entire story by any means, if you put that much L2 cache on an AMD it would make little or no difference, they are completely different architectures that require completely different parameters, AMD's architecture does not need the extra L2 where as Intels is enhanced by it, I suppose my point is, the extra L2 in itself does not make the gains clock for clock between the 2 architectures, it;s the efficiency within those architectures that determine the speed, for example, AMD is/always has been very strong on floating point calculations and memory intensive processes (AKA on Die mem controller), Intel is much stronger on arithmatic blah blah.......SceinceMark 2 sees the two of them more closely matched clock for clock as that bench tests memory and various CPU calculations.

Sorry long winded answer to a very small point!

One last point, you say, AMD's architecture is more advanced, that could be true, however, it means nothing if it cant compete, and sadly, across the board it cant compete.....now before you think it (if you havent already!) I was brought up on AMD, I love AMD and I will always go back to them.....when they become more competative.

Yea thats all true to, but i think having a bigger amount of Cache is a advantage, no matter what, like the old days, were 1MB of L2 cache was huge on a CPU, like mine < it gave a good boost in games mainly, but also a few other things. Im not worried about higher clock speeds, i just want to see a AMD vs a intel with the same amount of Cache and see what happens?at the same clock speeds, because at this time there is no comparison?
But agreed that AMD is a bit slow in a few other areas, AM2 was not a big leap from the old 939 at all =/ shame tho, just like to see the new 45nm ones with the CPU's of intel today, i think it might be interesting to see

 

[hat]

yeah AM2 was just 939 with DDR2 support

 

[Wile E]

Fudzilla said Deneb will be K11, I think?

I meant the next major arch change, not just a "tweak and shrink"™.

 

[Bluefox1115]

My argument against this comparison is the 12MB cache vs 2MB, which does make quite a large difference. Also, the current intel quads are 45nm, and AMD are 65nm, so forgive me in my lack of updating my brain as to the new 45nm vs 65nm previous gen. and nothing I've seen so far has come straight from AMD, so I went to AMD myself and checked out the press releases. the 45nm quad cores are set to start at 2.8, 3.0, 3.2 and so on. with a TDP rating UP TO 125 watts @ 3.2GHz.

Edit: also, according to AMD's site, the die shrink will in fact reduce power consumption and allow for better power management and usuage. This contradicts Xbit labs "prediction" as they have said themselves, merely a prediction, no supporting evidence.

 

[Melvis]

umm 5yrs? i doubt it.... give it mabey 8 months lol.... but who know's.... im AMD all the way... cheers mate.... off to bed...-out-

LOL yea ok yea, i hope these 45nm ones will do well, i think they will match it with what intel has now, just not sure for how long after that. Righto dude, catcha

 

[Melvis]

I really hope AMD does pull a rabbit out of their hat with Deneb. But overall, I think we may have to wait for K11 (or whatever the next architecture is called) for there to be a significant shift in performance.

Let it be known, I am not an Intel fanboy. I will switch back to AMD the instant they take the OCed performance lead. (Actually, I still have an AMD rig with a 6400+ in it. I just wish the damn board supported Phenom. )

yea it be nice to see AMD to match it with intel again, just for us, you know, bring the dam price down a bit , hello, those high end quads of intel are OMG expensive

What mobo you using with the 6400? O ris it only the AM2+ that take Quads?

 

[Wile E]

yea it be nice to see AMD to match it with intel again, just for us, you know, bring the dam price down a bit , hello, those high end quads of intel are OMG expensive

What mobo you using with the 6400? O ris it only the AM2+ that take Quads?

Some AM2 boards can take Quads, it depends on the size of the BIOS chip the manufacturer put on the board, and whether or not they choose to support it.

As for my board, it's a DFI NF UltraII M2. They aren't supporting Phenom on it. Don't know if it's a bios limitation, or their choice tho.

 

[Wile E]

My argument against this comparison is the 12MB cache vs 2MB, which does make quite a large difference. Also, the current intel quads are 45nm, and AMD are 65nm, so forgive me in my lack of updating my brain as to the new 45nm vs 65nm previous gen. and nothing I've seen so far has come straight from AMD, so I went to AMD myself and checked out the press releases. the 45nm quad cores are set to start at 2.8, 3.0, 3.2 and so on. with a TDP rating UP TO 125 watts @ 3.2GHz.

Edit: also, according to AMD's site, the die shrink will in fact reduce power consumption and allow for better power management and usuage. This contradicts Xbit labs "prediction" as they have said themselves, merely a prediction, no supporting evidence.

AMD has 2MB of L2 and 2MB of L3, for a total of 4MB. I still don't see cache being the deciding factor here.

 

[Melvis]

My argument against this comparison is the 12MB cache vs 2MB, which does make quite a large difference. Also, the current intel quads are 45nm, and AMD are 65nm, so forgive me in my lack of updating my brain as to the new 45nm vs 65nm previous gen. and nothing I've seen so far has come straight from AMD, so I went to AMD myself and checked out the press releases. the 45nm quad cores are set to start at 2.8, 3.0, 3.2 and so on. with a TDP rating UP TO 125 watts @ 3.2GHz.

Edit: also, according to AMD's site, the die shrink will in fact reduce power consumption and allow for better power management and usuage. This contradicts Xbit labs "prediction" as they have said themselves, merely a prediction, no supporting evidence.

YAY i have someone on my side LOL

 

[Bluefox1115]

its not shared L2, its dedicated native to each core hence 4x512k, unlike Intel. just the like the 9800gx2 is 2x512mb 2x256bit.

 

[Tatty_Two]

My argument against this comparison is the 12MB cache vs 2MB, which does make quite a large difference. Also, the current intel quads are 45nm, and AMD are 65nm, so forgive me in my lack of updating my brain as to the new 45nm vs 65nm previous gen. and nothing I've seen so far has come straight from AMD, so I went to AMD myself and checked out the press releases. the 45nm quad cores are set to start at 2.8, 3.0, 3.2 and so on. with a TDP rating UP TO 125 watts @ 3.2GHz.

Edit: also, according to AMD's site, the die shrink will in fact reduce power consumption and allow for better power management and usuage. This contradicts Xbit labs "prediction" as they have said themselves, merely a prediction, no supporting evidence.

Do you not think that AMD would bolt on extra L2 memory if it would bolster their performance to any degree?.....of course they would, it would pay for itself in a month because so many more people would buy AMD if they were competative....no, as I said, the architecture, unlike Intels would benefit little, AMD can swop out the L2 so much quicker than Intel so it dont need the huge amount's Intel does.

 

[Wile E]

its not shared L2, its dedicated native to each core hence 4x512k just the like the 9800gx2 is 2x512mb 2x256bit.

Well, Kentsfield doesn't have a fully shared L2 either, but it beats the Phenom in most test, and Phenom has the shared L3 to it's advantage vs those cpus.