AMD SOCKET AM2 ROADMAP SHAKEUP

9-inch

Distinguished
Feb 15, 2006
722
0
18,980
More info on upcoming socket AM2 processors.

AMD will release two single core and 7 dual core AM2 processors (without counting the energy efficient ones).

http://www.dailytech.com/article.aspx?newsid=850

The Energy Efficient CPUs will have a slightly redesigned Sleep, Power and Suspend states. These states are controlled by the operating system, but will require BIOS updates.


AMD Energy Efficient Desktop Models
Processor Socket Core Wattage Launch Date
Athlon 64 4800+ AM2 Dual 65W Q3'06
Athlon 64 4600+ AM2 Dual 65W 6/6/06
Athlon 64 4400+ 939 Dual 65W 6/6/06
Athlon 64 4200+ AM2 Dual 65W 6/6/06
Athlon 64 4000+ AM2 Dual 65W 6/6/06
Athlon 64 3800+ AM2 Dual 35W 6/6/06
Athlon 64 3800+ AM2 Dual 65W 6/6/06
Athlon 64 3500+ AM2 Single 35W 6/6/06
Sempron 3500+ AM2 Single 35W Q4'06
Sempron 3400+ AM2 Single 35W 6/6/06
Sempron 3200+ AM2 Single 35W 6/6/06
Sempron 3000+ AM2 Single 35W 6/6/06

65W X2 4800+ AM2 8O 8O 8O
That's really impressive. This is more proof that AMD's manufacturing process is doing really good. This one will be overclockers choice for sure.
 

Ycon

Distinguished
Feb 1, 2006
1,359
0
19,280
Those "energy efficient" CPUs still consume more than double the power that Yonah (also used for HTPCs) consumes.

And its a shame that AMDs Value CPUs (Sempron, Single Core!) will consume roughly the same as Intels top-of-the line Conroe (E6700, Dual Core!).
 

9-inch

Distinguished
Feb 15, 2006
722
0
18,980
Those "energy efficient" CPUs still consume more than double the power that Yonah (also used for HTPCs) consumes.

There will be a 35W X2 3800+. 8)

This is more proof that Dual core Turion are cappable of 2.6GHz and ONLY consumes 30-35W of power. :wink:

Keep trying harder fanboy.
 

Ycon

Distinguished
Feb 1, 2006
1,359
0
19,280
Those "energy efficient" CPUs still consume more than double the power that Yonah (also used for HTPCs) consumes.

There will be a 35W X2 3800+. 8)

This is more proof that Dual core Turion are cappable of 2.6GHz and ONLY consumes 30-35W of power. :wink:

Keep trying harder fanboy.
Even if there would be a 35W X2 3800+ (which I doubt cause it just doesnt fit the pattern) it still consumes more than Yonah and delivers worse performance.
 

9-inch

Distinguished
Feb 15, 2006
722
0
18,980
Even if there would be a 35W X2 3800+ (which I doubt cause it just doesnt fit the pattern) it still consumes more than Yonah and delivers worse performance.

Are you really sure about that?? :lol:

OK. according to this review (which many of you praise and love), the Core Duo consumes 92-108W. In notebooks, the damn thing consumes 49W, not even close to Intel's maximum (or average) TDP. :lol:

Here's another jewell:
http://everythingapple.blogspot.com/2005/10/yonah-is-power-hog.html

The "T" class of Yonah chips, which are expected to be fitted in most business notebooks, will come with a maximum power consumption of between 25 and 49 watts. Right now, single-core Pentium Ms top out at 27 watts.
If true, that's not entirely unexpected given that these chips have two complete and independent cores, but it is still disappointing. Historically, Apple has preferred CPUs with a max power utilization under 30 Watts. However, the specifications published by CNET indicate that dual-core 2+ GHz Yonah chips are far above that 30 Watt barrier.
:lol: :lol: :lol:
 

Ycon

Distinguished
Feb 1, 2006
1,359
0
19,280
Dont be an ignorant (I guess now its too late to say that)... Theres only a handful of NBs that have 100W PSUs...
 

9-inch

Distinguished
Feb 15, 2006
722
0
18,980
I think the ignorant is someone else who still thinks that a shitty Core Duo still consumes 30W (maybe the low-end versions). :lol: :lol:
 

CompGeek

Distinguished
Jan 23, 2005
455
0
18,780
Yonah consumes very little that's a fact. Of course there will always be a rubbish site as Cnet or Zdnet that may disagree here,but i don't take such sites seriously.
Everyone should know that.
X2 energy efficient? Not only that they still aren't as energy efficient as Yonah but they'll cost a lot more. AMDs main advantage,the price/perf , will no longer stand here. You'd be dumb to think that AMD is selling a X2 3800 at the same price as a tweaked X2 3800. The price difference will be somewhere above the n.p.p./4
where n.p.p.=normal product price
So the equivalent of a 600 $ normal CPU would be above 750
Intel doesn't need this,since they'll be offering high perf. at low energy cost no matter the product and as i saw,prices will be right.
 

ltcommander_data

Distinguished
Dec 16, 2004
997
0
18,980
This is more proof that AMD's manufacturing process is doing really good.
AMD's is continuing to refine their 90nm process, I'll give them that but that is not the primary reason for those Energy-Efficient processors. All you have to do is look at the recent FX-60 and soon to be released FX-62 to confirm that. The reason why the FX-60 fits into the power envelop of the X2 4800+ was that AMD lowered the voltage in order to constrain the thermal levels. This is simply binning a chip with a higher clcok speed to run at slower speeds to save power and is the same practice that is used to have low voltages in the Opteron line. For instance chips that may run stably at 2.8GHz at stock voltage of the X2s would be run at 2.6GHz in stable operation at lower voltages to reduce power consumption creating the FX-62. If the process was significantly removed AMD wouldn't have to result to those measures.

Similarly, if the process was so much improved there wouldn't be a reason to increase the TDP of the FX-62 to 125W. I'm sure people would argue that AMD is simply leaving extra room for higher-clock speeds. However, this argument is inheirently flawed. The next speed bump for the FX line will be the FX-64 which is a 65nm chip to launch in Q1 2007. Transitioning to 65nm for the next FX should lower the TDP, which means the increase is not for higher clock speeds but simply because the 90nm FX-62 genuinely needs it. This can hardly be proof of the 90nm process "doing really good".

http://www.channelregister.co.uk/2006/02/17/amd_energy_efficient_cpus/

The more likely reason that Energy-Efficient processors are possible is the one mentioned in the article. Specifically, the redesigned Sleep, Power, and Suspend states. It will be important to see if these processors are Energy-Efficient at the expense of some performance because of their more aggressive power-saving measures, which probably mean they downclocking more and deeper for a given system load.

It's also interesting that this article mentions nothing about actual AM2 clock speeds. This is especially important because of the controversy over whether AMD will keep the current PR naming structure for given clock speeds or increase it by 400MHz for AM2. The latter would of course mean that a X2 3800+ actually runs at 1.8GHz and the current 2GHz X2 3800+ becomes the X2 4200+. This would certainly explain how the X2 3800+ can be available at 35W using the 90nm process. The drastic cut from 89W to 35W truly seems "incredible" otherwise. Of course, this last paragraph is completely speculation on my part so you can discount it, but it seems reasonable considering that the AM2 processors were originally announced with the 2GHz model being the X2 4200+.

http://www.theinquirer.net/?article=28367
 

9-inch

Distinguished
Feb 15, 2006
722
0
18,980
Similarly, if the process was so much improved there wouldn't be a reason to increase the TDP of the FX-62 to 125W.
If you read the article I've posted about the FX-62 with 4MB L3 cache, then you'll know why the processor is heating that much (let's not forget that it will also use DDR2-800 memory controllers). I really hope this processor to break the FX-60 3.2GHz overclock on stock air cooling.

The more likely reason that Energy-Efficient processors are possible is the one mentioned in the article. Specifically, the redesigned Sleep, Power, and Suspend states. It will be important to see if these processors are Energy-Efficient at the expense of some performance because of their more aggressive power-saving measures, which probably mean they downclocking more and deeper for a given system load.

They'll use this processors for their Live platforms which will compete with Intel's ViiV since the concept of a media center is performance at lower power consumption.
 

ltcommander_data

Distinguished
Dec 16, 2004
997
0
18,980
If you've read the article on the 4MB L3 cache you'll know that the FX-62 doesn't have a L3 cache.

http://www.theinquirer.net/?article=29738

So, what is AMD going to do to counter the upcoming Conroe based XEs, not to mention Kentsfield? Well, the obvious answer is the FX62, and that will be no slouch.
What do you do when it is not enough? Cache.
Cache will be introduced on the FX-62's replacement.

AMD may wait until it sees competition before unleashing this, but it will happen sooner rather than later.
The increase in heat cannot be due to an L3 cache since the FX-62 doesn't have any. Even the DailyTech article that you started this thread with indicates that the 125W TDP is for FX processors with 2x1MB L2 cache, nothing about L3 cache is mentioned. You can see that in the picture on the right.

If you say that the Energy-Efficient processors are only for the Live platform that's not very competitive then. Intel's mainstream Conroes will have a 65W TDP and that's for desktop not just media centre use.
 

hergieburbur

Distinguished
Dec 19, 2005
1,907
0
19,780

There is a lot of sepculation in that article, though the concept is intriguing. I would think, however that is the L3 cache has lower latency than the L2, that they would eliminate the L2, otherwise they ahre hurting themselves. If not, that begs the question as to whether there are some stability data integrity problems with the Z-RAM that would make them steer away from using it as L2
 

Ycon

Distinguished
Feb 1, 2006
1,359
0
19,280
There could be another problem with the L3 Cache. The FX-62, already a hell of an expensive processor, could be simply too expensive with the Cache added. Dunno if AMD actually cares, especially looking at their pricing of the last few months, but that MIGHT lead to a problem with the chip.
 

Vic

Distinguished
Apr 18, 2004
156
0
18,680
Woh, Woh, woh Sweet Child of mine!

What some customers and PC builders have found is that single core applications have major problems with Dual Core set ups. This is massively manifested in lagging frames, even though the user has the best components and graphics card possible.

I like AMD - that's why I bought AMD. The Intel dual core offering is not clearly as well though out. But the main problem from all of this is the OS itself. XP seems to take dual core on the face of it, but when you get performance degradation due to lack of dual core support, its a tad disappointing!!!

Heres one MS article:

http://support.microsoft.com/Default.aspx?id=896256
 

TRENDING THREADS