Fuddo: AMD 45nm K10.5 scheduled to launch in 1H 2009

[yomamafor1]

http://www.fudzilla.com/index.php?option=com_content&task=view&id=5687&Itemid=1

It looks like AMD is also working on a new version of Tri-core and the company plans to run some interesting marketing for these parts. Tri-core will probably be very close to performance to any quad- core, especially in everyday applications and AMD wants to benefit that fact.

The new K10.5 is 45nm and it is codenamed Heka. It is AM2+ and AM3 compatible and we are expecting two different versions. The AM3 version supports DDR3, while the AM2 part supports DDR2 memory.

The new Tri-core chip will have 6MB of L3 cache and there will be a version without L3 cache, probably significantly cheaper.

This makes sure that K10-based Toliman is not just a one-time thing and that AMD plans to continue it in 2009.

Not that I have much faith in Fudzilla in terms of information accuracy, but it is a lot better than other speculative hardware sites (Inq anyone? ) If this is true, it certainly means AMD has yet to have 45nm K10 booted, contrary to what AMD claimed in the first place.

Then we'll be looking at a end of H1~H2 launch time frame for 45nm "Hekka".

 

[Kamrooz]

I'm so not surprised. 45NM in 2008 was impossible for AMD, I'm quite fed up with their lies and deceit. Is it so hard to actually give a more realistic timeline release? Instead of empty promises which end up with delays and hurts their image even more.

I was expecting 45nm from AMD in 2009, more likely late 1H or early/mid 2H 2009.

 

[Amiga500]

Is this tri-core for H1 2009 or quad-core for 2009?


The article wasn't explicit on that point I felt.

 

[Kari]

that was only about the tricore though, 45nm quads could still be on schedule for Q4 08

 

[cnumartyr]

Is this tri-core for H1 2009 or quad-core for 2009?


The article wasn't explicit on that point I felt.

Well since the Tri-cores are defective quads I would assume that the quads will come first... unless their yields are THAT bad.

 

[Amiga500]

Well since the Tri-cores are defective quads I would assume that the quads will come first... unless their yields are THAT bad.

Yeah - I understand that ok.


But was the article talking about quads or tris for H1 '09?

 

[yomamafor1]

The article is ambiguous itself. However, I speculate AMD to launch 45nm product in 2009, not 2008. They still have yet to come up with a boot silicon for 45nm K10, and it will take them approximately a year to iron out the issue, tune the process, and to bring them to market. (under ideal condition)

 

[cnumartyr]

The article is ambiguous itself. However, I speculate AMD to launch 45nm product in 2009, not 2008. They still have yet to come up with a boot silicon for 45nm K10, and it will take them approximately a year to iron out the issue, tune the process, and to bring them to market. (under ideal condition)

I don't think 45nm will do anything for them.

First generation will use SOI. They are still working out the bugs and problems with K10.

This just has me doing what I was thinking about doing all along... X48+DDR3+Q9450 in the Summer.. If AMD can't compete Intel isn't going to bother.

 

[yomamafor1]

I don't think 45nm will do anything for them.

First generation will use SOI. They are still working out the bugs and problems with K10.

This just has me doing what I was thinking about doing all along... X48+DDR3+Q9450 in the Summer.. If AMD can't compete Intel isn't going to bother.

AMD will have to wait to 2nd generation of 45nm to actually improve K10's performance by a substantial margin. 45nm SOI simply won't work well for them.

I would stay away from DDR3 for now , and use them as Nehalem hits.

 

[cnumartyr]

AMD will have to wait to 2nd generation of 45nm to actually improve K10's performance by a substantial margin. 45nm SOI simply won't work well for them.

I would stay away from DDR3 for now , and use them as Nehalem hits.

I just don't know yet. I have a feeling Nehalem will just get pushed back further and further.. Atleast until 2009.

Considering AMD won't have a 45nm Product out and K10 doesn't compete well with Core 2 and the 1st Generation SOI won't help... yea.

I have a feeling the Core2 will be good for 2-3 years.

 

[yomamafor1]

I just don't know yet. I have a feeling Nehalem will just get pushed back further and further.. Atleast until 2009.

Considering AMD won't have a 45nm Product out and K10 doesn't compete well with Core 2 and the 1st Generation SOI won't help... yea.

I have a feeling the Core2 will be good for 2-3 years.

Intel will probably release Nehalem for server first, then delay desktop product line for a while. With Nehalem in the server arena, I really don't see how AMD will survive without becoming... low end.

 

[cnumartyr]

Intel will probably release Nehalem for server first, then delay desktop product line for a while. With Nehalem in the server arena, I really don't see how AMD will survive without becoming... low end.

It's needed in the server market.

The FSB is still more than enough in the desktop market and will be for quite some time.

 

[yomamafor1]

It's needed in the server market.

The FSB is still more than enough in the desktop market and will be for quite some time.

Tell that to AMD

http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_15331_15332%5E15333,00.html


WOAH!!! Incredible performance!!

 

[Kari]

lol i guess you two have it all worked out lol

 

[cnumartyr]

Tell that to AMD

http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_15331_15332%5E15333,00.html


WOAH!!! Incredible performance!!

I have yet to be bogged down by a 3.6 GHz Q6600 with 1600 MHz FSB. Hell if I had DDR3-1600 the memory performance would be on par with the Phenom.

Got to love marketing gimics...

 

[yomamafor1]

I have yet to be bogged down by a 3.6 GHz Q6600 with 1600 MHz FSB. Hell if I had DDR3-1600 the memory performance would be on par with the Phenom.

Yes, but does it have INCREDIBLE PERFORMANCE?

 

[cnumartyr]

Yes, but does it have INCREDIBLE PERFORMANCE?

No, it has Phenominal performance...

 

[yomamafor1]

No, it has Phenominal performance...

It IS phenomenal... isn't it?

 

[cnumartyr]

It is phenomenal... isn't it?

I knew I spelled it wrong.. You know how sad it is to know that if K10 was just clock for clock as fast as Core2 this would be a completely different story?

 

[yomamafor1]

I knew I spelled it wrong.. You know how sad it is to know that if K10 was just clock for clock as fast as Core2 this would be a completely different story?

Nah, I was referring to Phenom's performance being..... Phenomenal .

If K10 is faster clock for clock, then yes, this would be a completely different story. People will be buying chips from both sides, and life will be good.

But since it didn't happen, we're all under the influence of the Dark Side.

 

[cnumartyr]

Nah, I was referring to Phenom's performance being..... Phenomenal .

If K10 is faster clock for clock, then yes, this would be a completely different story. People will be buying chips from both sides, and life will be good.

But since it didn't happen, we're all under the influence of the Dark Side.

That would have been the best scenario. If a 2.2 GHz Phenom performed the same as a 2.4 GHz Q6600. Whole new world then. Obviously AMD can't push the clocks as high as Intel (especially now at 45nm) so they really needed that advantage and didn't get it.

 

[massih]

3.Am2 was a big failure from AMD because: DDR performance was twice faster than DDR2 in 939 times! and If AMD did not switched today we had DDR1000(Perforemed like DDR3 2000!) Dual channel is a big gap between DDR2/3 and DDR.
user switched from good performed 939 to Intel Core2 .But not because they do not have any feature on their M/boards ,because they could not find 939 X2 CPUs easily ,then DDR ram went to fazed out.
Soooo AM2 is slower than 939!(per clock)

.Memoy controller in 9500 and other low-clocked phenoms is very poor,perform like a 754 DDR1 CPU,in clocks higher than 2.8 this problem solved,This means performance of phenom goes higher per clock! then a 3.2 GHZ phenom (shanghai) may beat a core2 per clock!who knows!

In intel Max of REAL performance of any Ram is this: 400MHZ for fsb800,533mhz for fsb 1066,667 mhz for fsb1333,800 mhz for fsb 1600, and this is a chipset limitation! and may change in next generation.

 

[Just_An_Engineer]

The article is ambiguous itself. However, I speculate AMD to launch 45nm product in 2009, not 2008.

The article appears to be referring to the socket AM3 processors, which were always scheduled for 2009 release. Everything I have read recently still has the 45nm AM2+ processors coming out later this year so we'll have to wait and see.

 

[epsilon84]

It looks like tri-core is here to stay. Personally, I think it adds too much segmentation to the market, but I guess AMD has to look after their bottom line...

 

[Mathos]

I don't know, most multithreaded apps right now don't scale well above 3 cores anyways.