Sign in with
Sign up | Sign in
Your question

"AMD’s New Processor Micro-Architecture In 2008"

Tags:
  • CPUs
  • Performance
  • Processors
Last response: in CPUs
Share
June 22, 2006 4:24:21 PM

Quote:
Advanced Micro Devices’ chief executive said that his company was planning to commercially release processors with new micro-architecture in 2008 after making pilot production in 2007. The introduction of the new micro-architecture may significantly boost performance of the company’s microprocessors, which are to face increased pressure from competing solutions by Intel Corp.

“We’re going to introduce a really new architecture that will work well with our partners for the best performance. We’re going to start sampling it at the end of 2007 and roll it out in 2008,” said Hector Ruiz, chief executive of AMD, in an interview with BusinessWeek.

“We’re in the throes of finalizing the architecture we’re going to introduce next, and that’s going to be killer,” Mr. Ruiz added.


http://www.xbitlabs.com/news/cpu/display/20060622014633.html

More about : amds processor micro architecture 2008

June 22, 2006 4:32:14 PM

Wow, 2008 was not what I wanted to hear. I have a feeling this will be the true answer to Conroe, not K8L, but we'll see.

P.S. Very interesting 9-inch, thanks for posting this.
June 22, 2006 4:33:24 PM

I hope they are under-promising in their stating new arch performance wil be better...

Quote:
Earlier it was projected that in 2007 the Sunnyvale, California-based chipmaker will introduce instruction extensions to the AMD64 micro-architecture and will bring-in HyperTransport technology in 2008. But recently the company said in a statement that AMD’s “next-generation architecture for servers, workstations and desktops” is planned to debut in mid-2007, and is expected to offer better platform performance-per-watt ratio as well as enterprise application performance.
Related resources
June 22, 2006 4:35:41 PM

Quote:
Wow, 2008 was not what I wanted to hear. I have a feeling this will be the true answer to Conroe, not K8L, but we'll see.

P.S. Very interesting 9-inch, thanks for posting this.


I guess K8L "IS" the true answer to conroe becasue of all the enhancements of the core itself, then we will see AMD moving to a new micro-architecture in 2008-2009 timeframe (which I guess will be K10).

Let's wait and see.

Quote:
P.S. Very interesting 9-inch, thanks for posting this.

You're welcome. :wink:
June 22, 2006 4:50:25 PM

Quote:
Wow, 2008 was not what I wanted to hear. I have a feeling this will be the true answer to Conroe, not K8L, but we'll see.

P.S. Very interesting 9-inch, thanks for posting this.


I don't think what Ruiz is talking about is K8L. But I could be wrong...

From what I've seen of K8L, it's not a major change in architecture.
June 22, 2006 4:54:51 PM

Quote:
I don't think what Ruiz is talking about is K8L. But I could be wrong...

From what I've seen of K8L, it's not a major change in architecture.


I agree, I wasn't clear enough. This new architecture in 2008 is probably going to be K10. However, inbetween K8 and K10 is K8L. I don't think K8L will be a Conroe killer, but I think K10 will be the answer to Conroe. I'm just disappointed at how far off that's going to be.
June 22, 2006 6:28:27 PM

The original article is here.

Quote:
Customers are finding the fact that we are opening up our architecture very attractive. There are some segments where raw chip performance is important, but that's a small segment of the market. As [Intel] closes the technology gap, it will be a much tighter race, but we're going to introduce a really new architecture that will work well with our partners for the best performance. We're going to start sampling it at the end of 2007 and roll it out in 2008.

Once again, we're going to distance ourselves from them. The good news for customers: [Intel is] getting better, but it's not a new architecture. We're in the throes of finalizing the architecture we're going to introduce next, and that's going to be killer.


And I think he is talking about K8L.

If he is not, then does he think that K8L is not good enough to talk about at this stage?
June 22, 2006 6:55:44 PM

I have but ONE problem. Why can't they have a K9? Don't know why, but it sounds so much cooler than K8L or K10. Kinda reminds me of Ken-L
June 22, 2006 7:29:14 PM

Problem: by 2008 Intel will have a new arch as well as a die shrink. If K8L doesnt compete well enough with Conroe then they will be more than a little late.
June 22, 2006 7:30:43 PM

LOL! K9: "not just for pet shops any more."
June 22, 2006 7:40:57 PM


ARE YOU SEEING THIS RUIZ! THIS IS TRUE INNOVATION! DARE TO DO THE IMPOSSIBLE! DARE TO SURPASS YOURSELF! Just use beef instead of horse.
June 22, 2006 7:44:11 PM

Quote:
LOL! K9: "not just for pet shops any more."
Yep. That processor would be a REAL dog. :wink:
June 22, 2006 7:48:16 PM

Let's hope it's not a b!tch to overclock...
Anonymous
a b à CPUs
June 22, 2006 8:28:24 PM

I think K9 is chillin with MPEG3...
I rember reading about K9 being droped quite a while ago, no reason given.

The way I see it, rationnaly, is that you have two team working, one of K9 and the other on K10, then you realise that the goal your trying to achieve with K9 are'nt possible/great, so you pull of the ressources, taking the nice stuff and assign them to K10.

this is PURE speculation and its based on what happen with MPEG3 vs MPEG4, mPeg 4 was aimed and low bandwith/mobile and they realise they could port it to Higher def/higher bandwith so it took over MPEG3.

But that might be the answer..
June 22, 2006 9:55:32 PM

There's no need to keep hounding the forums
June 22, 2006 10:02:07 PM

Your barking up the wrong tree... :lol: 
June 22, 2006 10:04:14 PM

:lol:  The AMD K9 would be great!!!
And a perfect companion to the IPod Flea!
:lol:  :lol:  :lol: 
June 22, 2006 10:08:55 PM

We're Sorry...
but we can't find the HomePage you've requested. It's possible that:
· The address was entered incorrectly. Check your spelling and try again.
· The .Mac member of this name has either created a page and removed it or has never published a HomePage.
· There is no .Mac member of this name. If you'd like this member name for yourself, sign up for a .Mac account right now and have your own HomePage in minutes.
June 22, 2006 10:13:06 PM

See? It's so invisible you can't even see the webpage. Now THAT is 1337.

Anyway, peek through the electron microscope here.
June 22, 2006 10:36:33 PM

Quote:
I guess K8L "IS" the true answer to conroe becasue of all the enhancements of the core itself, then we will see AMD moving to a new micro-architecture in 2008-2009 timeframe (which I guess will be K10).

Let's wait and see.


if it's the answer or not, intel will also be moving to a new architecture in 2008, which will also be on 45nm. if AMD can't match performance (or I guess more performance per watt) to core with K8L, then imo they HAVE to make a better architecture with "k10" than intels next core. if they can't, then at that point I think they may be in some trouble.

it's also funny how you're referring to "K8L" as an "answer" to conroe now. a few months ago, conroe performance numbers were all lies and you were sure AMD would pull an ace out of it's sleeve.
June 22, 2006 10:43:14 PM

He's actually stopped arguing / debating for now. Instead there is just a constant stream of pro-AMD press links each day, searching for the needle in the haystack.

Pity there is a lot of hay...
June 22, 2006 10:45:28 PM

That is what defines fanbois...
Commendable if 9nm acknowledges AMD has been dethroned... but I doubt he realized yet to what extent...
a c 487 à CPUs
June 22, 2006 11:01:10 PM

Intel has already annouced the successor to Conroe.

Intel Life After "Conroe" article.

Quote:

Penryn, which will be based on Intel's lithography process known as P1266, is a 45nm unified core set for launch in 2007 that is also expected to stay into production into 2008. Intel has already produced SRAM samples for 45nm CPUs, as demonstrated in the lithography shot on the right. Aside from the process shrink on Penryn, the major divergence in design from Conroe is the new material design. With P1266, Intel shifts away from Silicon Dioxide gate dielectrics -- a process the company has used since the mid-90s -- to High-k dielectrics. With new dielectric techniques, the company will also revamp its gate electrodes to metal instead of Polysilicon derivatives. The last major materials change of such magnitude occurred when Intel moved from Silicon to Strained Silicon in 2002, which is still slated for use in P1266 and beyond.


It seems you are incapable of providing both side of the story.
June 22, 2006 11:02:42 PM

which is due to his lacklusterly supported 1D vision...
June 22, 2006 11:05:06 PM

K9 microarchitecture may have been shelved to work on K10 for now. But it'll be back- it was just far too advanced for today! :p  Check again in the year 5000.




K9 info
Wikipedia entry for K9 <-- what's with this link not displaying??
June 22, 2006 11:06:04 PM

Could 1266 refer to the number of pins? That would be cool.
June 22, 2006 11:42:08 PM

Quote:
Could 1266 refer to the number of pins? That would be cool.


Just a internal code for the lithography process and materials used.
June 22, 2006 11:48:06 PM

Quote:
it's also funny how you're referring to "K8L" as an "answer" to conroe now. a few months ago, conroe performance numbers were all lies and you were sure AMD would pull an ace out of it's sleeve.


:lol: 
OK, let me see...

1) They will have an all-improved 65nm process which will increase processor performance to 20-25%.

2) They have 4x4 which will undoubtly be the "enthusiast platform" of choice.

3) They have Rev.G.

4) ...And of corse, they have K8L.

8)
June 22, 2006 11:55:45 PM

Quote:
it's also funny how you're referring to "K8L" as an "answer" to conroe now. a few months ago, conroe performance numbers were all lies and you were sure AMD would pull an ace out of it's sleeve.


:lol: 
OK, let me see...

1) They will have an all-improved 65nm process which will uncrease processor performance to 20-25%.

2) They have 4x4 which will undoubtly be the "enthusiast platform" of choice.

3) They have Rev.G.

4) ...And of corse, they have K8L.

8)

1) Speculation based on previous speed increases of the K8 uArch in contrast to nm process, with that in mind this "performance" will likely be power consumption, and thermal output improvements, clock speed increases almost don't seem feasable at this point.

2) I will go out and tell all my enthusiast friends they can buy 2 new 5100 Xeon's and run them on a 4x4... I mean DP workstation setup.

3) Oh the letter G that stands for Ganster that must mean the revision will be uber good in all aspects.

4) Core 3.
June 23, 2006 12:10:32 AM

Quote:
it's also funny how you're referring to "K8L" as an "answer" to conroe now. a few months ago, conroe performance numbers were all lies and you were sure AMD would pull an ace out of it's sleeve.


:lol: 
OK, let me see...

1) They will have an all-improved 65nm process which will increase processor performance to 20-25%.

2) They have 4x4 which will undoubtly be the "enthusiast platform" of choice.

3) They have Rev.G.

4) ...And of corse, they have K8L.

8)

1. speak all you want about AMD's 65nm, but the only real data we've heard about it is straight from AMD. and if I remember correctly, according to you, any data given by the manufacturer is "bs", and we have to wait for real world samples to judge the tech by. .. at least that's what you said about intel's info. don't be a hypocrite.

2. 4x4 at this point is, in all fairness, ridiculous for at least the gamer enthusiast. this _has_ to put some doubt in even the most biased AMD fans when it comes to AMD's "pure" intentions when they market a 4 core platform as "top of the line" for gamers. everyone knows games hardly take advantage of 2 cores, let alone 4. and also, look at all the slack intel got for pushing their EE with HT for gamers. 4 threads for gamers, same situation; unecessary.

other enthusiasts might go for it, like the heavy duty multitaskers or video editors and whatnot. for them it makes sense. but realistically, the majority of enthusiasts are going to go for the fastest dual core cpu.

3 and 4. again not enough info to make an argument out of.
June 23, 2006 12:12:44 AM

Quote:

1) Speculation based on previous speed increases of the K8 uArch in contrast to nm process, with that in mind this "performance" will likely be power consumption, and thermal output improvements, clock speed increases almost don't seem feasable at this point.

Uhmm.. Are you forgetting embedded SiGe with dual stress liner and stress memorization technology on silicon on insulator?? :wink:

Quote:
I will go out and tell all my enthusiast friends they can buy 2 new 5100 Xeon's and run them on a 4x4... I mean DP workstation setup

So, are you finally admitting that 4x4 will beat any conroe config call it dual or quad?

Quote:
3) Oh the letter G that stands for Ganster that must mean the revision will be uber good in all aspects.

:roll:

Quote:
Core 3.

K10. 8)
June 23, 2006 12:20:39 AM

Quote:
1. speak all you want about AMD's 65nm, but the only real data we've heard about it is straight from AMD. and if I remember correctly, according to you, any data given by the manufacturer is "bs", and we have to wait for real world samples to judge the tech by. .. at least that's what you said about intel's info. don't be a hypocrite.

I am not a hypocrite (as you state), is just that I'm a great witness of intel's BS and way to spread FUD (10GHz anyone??).

Quote:
2. 4x4 at this point is, in all fairness, ridiculous for at least the gamer enthusiast. this _has_ to put some doubt in even the most biased AMD fans when it comes to AMD's "pure" intentions when they market a 4 core platform as "top of the line" for gamers. everyone knows games hardly take advantage of 2 cores, let alone 4. and also, look at all the slack intel got for pushing their EE with HT for gamers. 4 threads for gamers, same situation; unecessary.

Still, it will be a killer platform that not even intel can touch with conroe.
Let's hope they leave it open for X2 processors also.

Quote:
3 and 4. again not enough info to make an argument out of.

I won't argue either so let's call it a tie. :wink:
June 23, 2006 12:27:56 AM

Quote:
I am not a hypocrite (as you state), is just that I'm a great witness of intel's BS and way to spread FUD (10GHz anyone??).


Oh the irony. Anyways how is that FUD? Their plan bombed.

Quote:
Still, it will be a killer platform that not even intel can touch with conroe.


Quote:
Let's hope they leave it open for X2 processors also.


:lol:  :lol: 
June 23, 2006 12:29:18 AM

Quote:

1) Speculation based on previous speed increases of the K8 uArch in contrast to nm process, with that in mind this "performance" will likely be power consumption, and thermal output improvements, clock speed increases almost don't seem feasable at this point.

Uhmm.. Are you forgetting embedded SiGe with dual stress liner and stress memorization technology on silicon on insulator?? :wink:

Quote:
I will go out and tell all my enthusiast friends they can buy 2 new 5100 Xeon's and run them on a 4x4... I mean DP workstation setup

So, are you finally admitting that 4x4 will beat any conroe config call it dual or quad?

Quote:
3) Oh the letter G that stands for Ganster that must mean the revision will be uber good in all aspects.

:roll:

Quote:
Core 3.

K10. 8)

1) Nope but from what I have seen they don't do a lick for AMD other than in thermal output and power draw, no where has AMD nor IBM for that matter said 100% no doubt I swear on my grandmothers grave that SOI will increase clock speeds, let alone the materials used in their lithography process. It's all uArch at this point for AMD 65nm will bring their thermal output and electrical draw down but a noticeable increase in clock speed I personally have trouble accepting.

X2 K8-65nm for sure 2600 and 2800 respectively 3000 I find a bit far fetched at this point, but I don't know what AMD has all in store for the process shift either they could very well surprise me.

2) They aren't even in the same playing field 1 dual core processor getting outpaced by 2 dual core processors, pretty feasible but then there is the Kentsfield.

3)Moo.

4) Core 4
June 23, 2006 12:34:35 AM

Quote:
I am not a hypocrite (as you state), is just that I'm a great witness of intel's BS and way to spread FUD (10GHz anyone??).


hey, AMD made that "40% transistor performance increase" statement. and you know they were trying to make their new manufacturing process sound 40% faster with that. all manufacturers will make themselves look as good as they can, even if it means skewing the facts a little.

you're a hypocrite if you believe what AMD says but not what intel says. AMD is here to make money, remember? they're not your best buddy that's "looking out for you". they're a company just like intel.

Quote:
Still, it will be a killer platform that not even intel can touch with conroe. Let's hope they leave it open for X2 processors also.


not killer for everyone, which is what I'm saying. it can't be a "conroe killer" if it's a high priced item geared towards a niche market.

intel's 4-core cpu will be a contender, but the 4x4 platform has the advantage of HT for sure. however, a 4-core intel cpu setup will probably be cheaper than a 2 dual core AMD setup. can't say for sure though.
June 23, 2006 12:42:18 AM

Quote:
1) Nope but from what I have seen they don't do a lick for AMD other than in thermal output and power draw...

So you're saying that venice isn't better than Clawhammer (performance/watt) wyse?
http://www.lostcircuits.com/cpu/amd_venice/[/quote]

Quote:
..., no where has AMD nor IBM for that matter said 100% no doubt I swear on my grandmothers grave that SOI will increase clock speeds, let alone the materials used in their lithography process. It's all uArch at this point for AMD 65nm will bring their thermal output and electrical draw down but a noticeable increase in clock speed I personally have trouble accepting.

Have a little more respect for your grandmother. :x
June 23, 2006 12:43:46 AM

Quote:

Still, it will be a killer platform that not even intel can touch with conroe.




So, your argument is AMD will have something out in 2008 that will beat Conroe. Its 2006 now. 2 years.

Woo whoo, I'll bet Intel is quaking in it's boots. This fast paced living AMD is doing will kill people...........of old age
June 23, 2006 12:46:51 AM

Quote:
1) Nope but from what I have seen they don't do a lick for AMD other than in thermal output and power draw...

So you're saying that venice isn't better than Clawhammer (performance/watt) wyse?
http://www.lostcircuits.com/cpu/amd_venice/

Quote:
..., no where has AMD nor IBM for that matter said 100% no doubt I swear on my grandmothers grave that SOI will increase clock speeds, let alone the materials used in their lithography process. It's all uArch at this point for AMD 65nm will bring their thermal output and electrical draw down but a noticeable increase in clock speed I personally have trouble accepting.

Have a little more respect for your grandmother. :x[/quote]

Thats what I am saying, it does help with thermal output and electrical draw.
June 23, 2006 1:29:55 AM

4x4 would be amazing if it was 2x2, as in two single cores... thats alot less overhead than 1 dual core... ask any server builder.... if am2 had a really good single core ie fx 57 and you could 2x2 them that would make one beafy rig.

but then again dual core clock speeds are up to single core ones now so my entire post might have been a waste >< sorry

edit. you would still have multi threaded proformance but at this time there is nothing in the gaming world optimized for 4 threads.
June 23, 2006 1:31:47 AM

9 Inch, I believe 4x4 is not AMD's answer to Conroe, but to Kentsfield. I don't think AMD's quad cores will be available in time for the release of Kentsfield so they had to improvise. AMD's quads are needed for their high-end server business so it will be a while before they arrive on the desktop. 4x4 will have to suffice until AMD can produce quads in large enough numbers.

I've used both Intel and AMD products throughout the years and I don't have a favorite one way or the other. My performance sweet spot is modest; 3 GHz on P4, 2 GHz on A64 or PM. What I will demand from my next CPU is low power consumption and reduced heat. But even if AMD lowers the price on their X2s to Conroe levels, I would still have to go with Intel because their architecture is brand new while K8 is at the end of its line.

I realize this is an enthusiast board and it saddens me that AMD is having to reallocate its resources to compete at the high end, with servers. They will obviously still be a force to contend with in the desktop and notebook markets, but they will be at a disadvantage for the next year or so. I would much rather have AMD remain viable, even if it means a temporary retreat in those two segments. If Opteron is what will keep AMD profitable, then that's where they should concentrate. Besides, enthusiasts can use Opteron (but they can also use Woodcrest).

Intel survived for 3 years despite being at a performance disadvantage (in desktops and servers, but not in notebooks). Now AMD must do the same, and count on their well deserved reputation to pull them through. Hopefully the disadvantage will only last 12-18 months. I wouldn't count on 65 nm to improve the K8's performance, but it will make it cheaper to manufacture. That could save AMD from a bath of red ink while allowing it to maintain market share.

Contrary to what you believe, an AMD/ATI merger would have been a huge coup for AMD. Core + Crossfire will be unbeatable, at least according to Charlie Demerjian, one of your favorite sources. AMD could've stopped that onslaught by buying ATI. Too bad it didn't happen. Now there's talk about SLI becoming available on Intel chipsets. That is very bad news for AMD.
June 23, 2006 2:08:47 AM

What matters here is to understand that "K8L" as it is right now, is not AMD's response to the Conroe. The truth is that AMD did not re invent their architecture, they simply expanded to a DDR2 compatible platform. Now, AMD's true response is on its way. Whether its their forecasted 65nm cores at the end of this year, or something we don't know about. Conroe is a new core on a relatively "old" platform, so, just like the most recent Athlon 64 cores, San Diego was an improvement over on ╟┼─┼╢[/img]
June 23, 2006 10:23:53 AM

Quote:
1. speak all you want about AMD's 65nm, but the only real data we've heard about it is straight from AMD. and if I remember correctly, according to you, any data given by the manufacturer is "bs", and we have to wait for real world samples to judge the tech by. .. at least that's what you said about intel's info. don't be a hypocrite.

I am not a hypocrite (as you state), is just that I'm a great witness of intel's BS and way to spread FUD (10GHz anyone??).

Quote:
2. 4x4 at this point is, in all fairness, ridiculous for at least the gamer enthusiast. this _has_ to put some doubt in even the most biased AMD fans when it comes to AMD's "pure" intentions when they market a 4 "core platform as "top of the line" for gamers. everyone knows games hardly take advantage of 2 cores, let alone 4. and also, look at all the slack intel got for pushing their EE with HT for gamers. 4 threads for gamers, same situation; unecessary.

Still, it will be a killer platform that not even intel can touch with conroe.
Let's hope they leave it open for X2 processors also.

Quote:
3 and 4. again not enough info to make an argument out of.

I won't argue either so let's call it a tie. :wink:


"I am not a hypocrite (as you state), is just that I'm a great witness of intel's BS and way to spread FUD (10GHz anyone??)."

So you beleive one company over the other based apon the fact its a peticular company..... that my friends is the definition of a fanboy.
June 23, 2006 11:11:45 AM

Quote:
Quote:
I will go out and tell all my enthusiast friends they can buy 2 new 5100 Xeon's and run them on a 4x4... I mean DP workstation setup

So, are you finally admitting that 4x4 will beat any conroe config call it dual or quad?


Wow man, you are making yourself looking dumber as you continue to post. The reason why he said to use Xeon platform because Intel never say that they will release a platform for 2 or more conroes. Hence, the conroe config for dual or quad will prolly never happen. You can't beat something that is not exist.
June 23, 2006 11:45:20 AM

Quote:

Still, it will be a killer platform that not even intel can touch with conroe.




So, your argument is AMD will have something out in 2008 that will beat Conroe. Its 2006 now. 2 years.

Woo whoo, I'll bet Intel is quaking in it's boots. This fast paced living AMD is doing will kill people...........of old age

:)  Do we expect 2 years from now Conroe to be the Conroe of today? :)  ... Actually, I think I read somewhere, yeah it will be Nehalem. So it is at this point ? vs ? is it not :) 
Why is it I read this artical a bit differantly than others. Knowing that Socket F will arrive at that very time gives great insight. The socket F is the first board to handle HT 3.0 so Im thinking this is Full Reverse HyperThreading the CEO is speaking of. I dont think FPU extensions, partial Reverse HyperThreading, will require HT 3.0 which is the plan to slowly step into the process.
June 23, 2006 12:36:24 PM

I think we're all going to have to get used to the fact that in the future, neither AMD or Intel are going to conclusively "beat" the other. Since Prescott the CPU market has been unhealthy, since the A64 has conclusively beaten the P4 in most things since then. As a result, it seems that the CPU market has been somewhat stagnant.

Yes, I imagine that Intel and AMD are going to switch positions, with Intel being the clear performance leader for the rest of this year. The best scenario for us as consumers is for K8L to be available H1 07 and to match Conroe. Similarly, AMD really needs K10 to be out in time to match up against Nehalem. That way we will have a real choice again that we havn't had for a long time, where we are able to pick CPU's based on their specific strengths for the kind of tasks we use. Look at ATI and Nvidia - it's rare for either one of them to have a clear across the board advantage for more than a month or two. Competition drives innovation
June 23, 2006 7:16:50 PM

Quote:
What matters here is to understand that "K8L" as it is right now, is not AMD's response to the Conroe. The truth is that AMD did not re invent their architecture, they simply expanded to a DDR2 compatible platform. Now, AMD's true response is on its way. Whether its their forecasted 65nm cores at the end of this year, or something we don't know about. Conroe is a new core on a relatively "old" platform, so, just like the most recent Athlon 64 cores, San Diego was an improvement over on ╟┼─┼╢
[/img]

4 some reason i cant edit my post, and it was posted b4 it was done :( 
so here im goig to complete it.

What matters here is to understand that "K8L" as it is right now, is not AMD's response to the Conroe. The truth is that AMD did not re invent their architecture, they simply expanded to a DDR2 compatible platform. Now, AMD's true response is on its way. Whether its their forecasted 65nm cores at the end of this year, or something we don't know about. Conroe is a new core on a relatively "old" platform, so, just like the most recent Athlon 64 cores, San Diego was an improvement over Clawhammer, as Venice was over Winnie. Conroe is probably going to be the leader of the pack until later this year. But since we don't yet have numbers from a comercialized product from this line, and as how the only numbers we do hae are those from intel's press version of the core. And we all know how intel likes tilting tables in their favor, and i have doubt if the version intel shipped out was actually a production model. And too bad TH compared it with the flaky FX62 instead of the solid FX60. And even if its better than AMDs. I doubt AMD would let themselves be beat for more than a short while. They prolly hard at work in the kitchen cooking up a new beast, which will be made available at a lower price than the conroes, as this is what AMD does so well. Ok, this post is long, and, well, biased. If i am wrong anywhere please quote the mistake and correct it.
June 23, 2006 7:53:28 PM

"which will be made available at a lower price than the conroes, as this is what AMD does so well."

You know this for sure? Intel has MUCH higher production capacity as well as lower costs for production, to stay competative it is likely they will keep prices low. Besides for intel they can still make a profit on their products at low prices, AMD cannot do the same. And no company has "skills" in keeping prices low, ether you do or you dont, making money while doing it however, is a different story.
June 24, 2006 1:22:26 AM

Quote:
Here are a few other FUD examples:
http://images.dailytech.com/nimage/1882_large_optyF.png

AMD's "estimate" for total power is completely off base here and is contradicts the measured values to date by a significant amount. If you read the comments they did not fool many people with this one.


So. you're calling this one a FUD???
I believe AMD is being kind with intel with these numbers but anyhow, let's do some math...
It's well known that FB-DIMM consumes 5W per module. Suppose you have a Woodcrest config with ALL your modules populated (let's say 16).
Multiply 16x5 and that will give you a total of 80 f*cking watts just for FB-DIMM alone.

I do hope AMD won't use FB-DIMM in their future incarnations becasue then they will run in the same problem intel is.

So, once again, who's the winner here? :wink:

Ohh, and talking about FUD, wasn't intel the one who said that their conrunt was 40% better than K8? :lol: 
      • 1 / 3
      • 2
      • 3
      • Newest
!