backtop


Print 61 comment(s) - last by TimberJon.. on Jul 31 at 4:18 PM


Slide previewing "Sandtiger" and third-generation Opteron platform  (Source: AMD)
New eight-core "Sandtiger" set for 2009 on third-generation Opteron platform

AMD previewed its third-generation Opteron platform during the company’s Technology Analyst. The third-generation AMD Opteron platform is set for a 2009 debut with a new Sandtiger-core processor. AMD’s current Socket 1207 platform is the Opteron’s second-generation platform and designed to accommodate Opteron revision F, Barcelona and the 45nm die-shrink Shanghai processors.

Sandtiger ups the core ante with eight-cores. AMD will manufacture the octal-core Sandtiger on a 45nm fabrication process. Sandtiger introduces HyperTransport 3.0, or HT3, to AMD’s server and workstation platforms. The processor will have four HyperTransport links with a new Direct Connect 2 infrastructure.

Sandtiger will also support DDR3 memory. The third-generation AMD Opteron platform will be the first platform to support G3 Memory Extender, or G3MX, technology for larger memory capacities. The platform also supports PCI Express 2.0.

New server and workstation chipsets will accommodate the new platform. AMD makes no mention on socket compatibility of the third-generation Opteron platform, but AMD’s track record shows socket changes with a new platform.

Before Sandtiger, AMD has Shanghai up its sleeves. Shanghai will be the 45nm shrink of Barcelona. AMD has Shanghai on the roadmap for a 2008 launch to replace Barcelona. Shanghai will drop into existing second-generation AMD Opteron platforms.

Shanghai features minor improvements such as a 6MB L3 cache and IPC improvements. The core will continue to have 512KB of L2 cache per core with three HT1 links. Registered DDR2 memory continues to be the memory of choice for Shanghai.

Expect AMD to unveil its third-generation AMD Opteron platform with G3MX technology and Sandtiger core in 2009 and Shanghai around the second half of  2008.


Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Sweet!
By Duraz0rz on 7/26/2007 2:09:12 PM , Rating: 4
Awesome stuff from AMD recently. Although Intel has the upper hand as of now, AMD is developing great stuff and building up on it. If Barcelona does well, I think the future is going to look bright for everyone.




RE: Sweet!
By webdawg77 on 7/26/2007 2:13:42 PM , Rating: 2
That's a BIG if . And no, I'm not an Intel fanboi. I have 3 X2 3800s and 1 X2 4200 running at the house.


RE: Sweet!
By helios220 on 7/26/2007 2:27:44 PM , Rating: 5
Regardless of personal preference, success of future AMD platforms will ultimately be most beneficial to the consumer. Intel has really done well recently, but for a long time before that we saw AMD preaching that innovative architecture was the key to future development while Intel just kept upping the clock frequencies on existing 32-bit platforms.

Regardless of whether you love Intel enough to name your child Pentium, solid competition from AMD has been a driving force for Intel to innovate. If AMD withers away again ultimately it’s going to result in less innovative product lines for everyone.


RE: Sweet!
By Master Kenobi (blog) on 7/26/07, Rating: 0
RE: Sweet!
By KristopherKubicki (blog) on 7/26/2007 3:01:13 PM , Rating: 3
To AMD's credit, I think 45nm is going to happen a lot faster than its 65nm ramp. The company put too many eggs in one basket with that ramp, but Fab 36 is not the only plans for 2008.


RE: Sweet!
By AMDfreak on 7/26/2007 3:08:21 PM , Rating: 5
quote:
#1, Intel has had 64-bit longer than AMD. (See Itanium)


This is true, but it should also be noted that AMD was the first to do 64-bit on x86. It is also notable that Intel copied AMD's 64-bit x86 solution.

AMD is having trouble with getting Barcelona's to run fast, and as a company seem to be floundering a bit. I hope they come back stronger than ever, but i don't see a light at the end of the tunnel yet.

As for me, I have succumbed to the Dark Side and have a G0 Q6600 on the way...


RE: Sweet!
By KTLA on 7/26/2007 3:09:17 PM , Rating: 2
quote:
#1, Intel has had 64-bit longer than AMD. (See Itanium)

There's no misconception. Everyone knows that Intel developed IA64 before AMD invented AMD64.

Everyone also knows that AMD64 won that fight, despite being released much later.

One may argue that IA64 is "better" that AMD64, but one CANNOT argue that AMD was the company that moved 64-bit from a miniscule fraction of chips to an everyday feature.


RE: Sweet!
By FITCamaro on 7/26/2007 3:18:08 PM , Rating: 3
IA64 does not have anything to do with x86-64. It is a true 64 bit design that can not run 32-bit code. Now of course though Intel has added backwards compatibility with 32-bit code but that originally was not the case.

x86-64 is the complete opposite being an extension of the x86 architecture that allows 64-bit code to be run.

IA64 was nothing new as true 64-bit chips were already in existence. It was just Intel's first try at it. It is also not even in the same category being mainly used for high end mainframes and supercomputers and cost over $1000 even for the lowest end models.


RE: Sweet!
By S3anister on 7/27/2007 5:27:02 PM , Rating: 2
NINTENDO SIXTY FOURRRRRRRRRRRRRRRRRRRRR!!!!!


RE: Sweet!
By miker75 on 7/29/2007 7:27:33 PM , Rating: 2
quote:
It was just Intel's first try at it. It is also not even in the same category being mainly used for high end mainframes and supercomputers and cost over $1000 even for the lowest end models.


True, but that wasn't Intel's original plan..
I remember reading years ago (when I was excited about the Itanium), that Intel was planning first to implement the Itanium on the server end, then to deploy cut-down versions into desktops and workstations...

But then as we all know, that changed when AMD brought out AMD64..


RE: Sweet!
By jconan on 7/31/2007 2:03:08 AM , Rating: 2
Before IA64 Intel did the same thing with IA32 x86-32 on Pentium II. PII was not very compatible with 16 bit applications and ran 16 bit code slow it did not sell very well. Hence Pentium III was born and as for the IA64 Intel probably forgot about the lessons from PII and it may be one of the reasons of Itanium's slow adoption into the marketplace and why x86-64 is successful.


RE: Sweet!
By Chadder007 on 7/26/2007 3:52:43 PM , Rating: 2
I just wish that companies would take more advantage of x64 than they have been. Hasn't there been only like 2 games that take advantage of anything a 64bit OS has to offer? Also, Adobe is going to be a few more years also in releasing an x64 suite of its products.


RE: Sweet!
By FITCamaro on 7/26/2007 3:09:40 PM , Rating: 2
They may have had a true-64 bit chip before AMD, but AMD was the first to have an x86 compatible 64-bit chip on the market which is what was needed to bring 64-bit computing to mainstream. Its not there yet really but if not for AMD pushing it, it likely wouldn't even be where it is now. Intel only added 64-bit extensions to compete with AMD, not the other way around.

And Sun had 64-bit chips over a decade before Intel. Not to mention IBM.


RE: Sweet!
By TomZ on 7/26/2007 10:28:43 PM , Rating: 2
No, the real reason that 64-bit has such small market share is that, on the desktop, there is no need for it today for most people. There will be in the near future, and there is need today in servers and HPC, but for desktop, there's not much need for it. I realize there are exceptions to that rule, but I'm talking about the 95% of desktop computers out there - there's no benefit to 64-bit.

AMD created AMD64 as a marketing diffentiator, one that paid off. They created something pretty far ahead of market need, which was pretty smart, although it was technically unnecessary at the time. Kudos to AMD for that.


RE: Sweet!
By Calin on 7/27/2007 4:09:02 AM , Rating: 2
People bought the AMD64 based on performance, not only on 64-bitness. True, in Enterprise market the Opteron was faster than Netburst Xeons, used less power and the 64-bit was put to use. However, now as games start to push the 32-bit addressable memory, transition to 64bit should happen soon at high end and invade the lower end also


RE: Sweet!
By mdogs444 on 7/26/07, Rating: 0
RE: Sweet!
By omnicronx on 7/26/07, Rating: 0
RE: Sweet!
By evildorf on 7/26/2007 5:05:13 PM , Rating: 4
quote:
it will probably take intel longer to perfect their manufacturing process below 45nm, just because they lack AMD/IBMS soi process, amd could play some catch up here.


You always to seem to bring up SOI in these discussions. The technology is widespread and by no means exclusive to AMD/IBM. Intel simply hasn't incorporated it because it adds significantly to manufacturing costs. If they saw a need for it, they would use it, and it's my understanding that it is much more easily implemented (cost excluded) than many other process technology modifications. This isn't to say that AMD/IBM aren't going to catch up, it just most likely won't be because of SOI.


RE: Sweet!
By TomZ on 7/26/2007 10:48:15 PM , Rating: 2
I also think I read somewhere that SOI isn't useful below 65nm anyway. I could be wrong about that - anybody know for sure?


RE: Sweet!
By Schmeh on 7/26/2007 3:47:55 PM , Rating: 3
quote:
Wrong.

During much of its history, AMD has developed and produced processors patterned after Intel's original designs, but in a twist of computing history, Intel found itself in the position of adopting AMD's x86-64 architecture, which AMD had created as an extension to Intel's own x86 processor line. As AMD later renamed its architecture to AMD64 (which includes Athlon, Opteron, Sempron, etc), Intel in turn has renamed their implementation Intel 64. It is used in newer versions of Pentium 4, Pentium D, Pentium Extreme Edition, Celeron D, Xeon, and Pentium Dual-Core processors, and in all versions of the Core 2 processors.


Actually he is not wrong. Read what he said again:

quote:
#1, Intel has had 64-bit longer than AMD. (See Itanium)


Notice he said nothing about x86-64, and he referenced Itanium. Itanium was released in 2001, while Opteron (K8) was released in 2003. So technically Intel had a 64-bit processor 2 years before AMD.


RE: Sweet!
By Amaroque on 7/29/2007 2:50:45 PM , Rating: 2
quote:
During much of its history, AMD has developed and produced processors patterned after Intel's original designs, but in a twist of computing history, Intel found itself in the position of adopting AMD's x86-64 architecture, which AMD had created as an extension to Intel's own x86 processor line. As AMD later renamed its architecture to AMD64 (which includes Athlon, Opteron, Sempron, etc), Intel in turn has renamed their implementation Intel 64. It is used in newer versions of Pentium 4, Pentium D, Pentium Extreme Edition, Celeron D, Xeon, and Pentium Dual-Core processors, and in all versions of the Core 2 processors.


Intel copied AMD's 3D Now! FP SIMD (vectorized) extensions idea on X86 with SSE. The K6-2 was the first CPU to introduce a floating point SIMD instruction set (dubbed 3DNow! by AMD). Intel improved upon it, but AMD innovated it.


RE: Sweet!
By Schmeh on 7/26/2007 3:41:51 PM , Rating: 3
quote:
#4, The only variation from the original Pentium architecture was P4, which flopped.


I would strongly dispute this statement. Are you forgetting about the P6, the basis for the Pentium Pro, Pentium II, Pentium III, and Pentium M. The P6 introduced speculative execution and out-of-order completion, register renaming, an integrated L2 cache, as well as other improvements and design changes.


RE: Sweet!
By Master Kenobi (blog) on 7/26/07, Rating: -1
RE: Sweet!
By Schmeh on 7/26/2007 5:39:37 PM , Rating: 2
quote:
It was an evolution of the original Pentium architecture. Pentium 4 was a vast departure from previous generations of chips.


I can possibly see why you would say that the P6 was a evolution of the P5. But I don't see you think Netburst was a vast departure from the previous designs. Unless you are talking about the length of Netburst's pipeline compared to the Pentium 3's.


RE: Sweet!
By emboss on 7/27/2007 1:54:54 PM , Rating: 2
The P6 architecture had nothing to do with the original Pentium architecture (which was really just an improved 486). It was a complete redesign with a RISC-like core, speculative & out-of-order execution, and a much deeper pipeline. It was designed to be very fast at 32-bit code, which is why the Pentium Pros were somewhat lackluster when it came to running 16 bit code (Win95 for example). I'd say that the differences between P5 and P6 were just as great as between P6 and P7.


RE: Sweet!
By emboss on 7/27/2007 2:02:09 PM , Rating: 2
Hmm, in future I should double-check against the parent post of the parent post to avoid repeating things ... :)


RE: Sweet!
By Chadder007 on 7/26/2007 4:01:19 PM , Rating: 2
Wasn't it IBM that came out with the first full 64bit processor?


RE: Sweet!
By dgingeri on 7/26/2007 4:19:14 PM , Rating: 2
quote:
Wasn't it IBM that came out with the first full 64bit processor?


Nope, actually, the first 64-bit processor was even earlier. It would be the MIPS V architecture. NEC, Toshiba, and SiByte were the first to introduce chips based on this in 1991. The first physical chip was the R4000 at 100Mhz.

MIPS also had hyperthreading before Intel.

http://en.wikipedia.org/wiki/MIPS_architecture

have fun reading about how Intel is the slow-show and RISC does, and always has, totally outstripped Intel in every way shape and form.

There is a reason there are currently fewer than 500,000 IA64 systems up and running in the world.


RE: Sweet!
By Topweasel on 7/28/07, Rating: 0
RE: Sweet!
By Hawkido on 7/30/2007 4:37:12 PM , Rating: 1
Yes but let's look at it this way:

#1: Itanic sucked, sucks, and will suck (plus a has been being suck for good measure) in everything except custom scientific apps.

#2: Motorolla 8086

#3: If you violate someone elses IP it usually doesn't result in a judge saying "okay you can use their IP" without some fishy going ons on the part of the original IP holder's part.

#4: Yet Intel clung to that turd and shoved it down our throats till AMD gave us a truely better option despite Intel's "questionable" business tactics trying to bury the competition.

#5: Yes, but Intel would be making MUCH more money (with less passed on to the customer I might add) at 130NM right now if they didn't have to keep ahead of AMD. Damn them!

#6: Why would there be a price cut if there was no AMD? Why bother sinking so much cash into a die shrink so soon after 65NM if you are still paying for the 65NM move? AMD.

#7: AMD's track record hasn't always been the best, most of that can be tracked to a small cash flow. The small cash flow can be traced to anti-competetive operations by Intel. And yet Intel is still sprinting with it's manufacturing and price cuts, because it now knows it can never stop pressing the line because as far behind as you say AMD is, Intel still feels what it was like to sniff AMD's butt back in the Pre-C2D days.

#8: Once I see a working processor at 32NM, then I will believe it. Remember there is a definite limit as to how small you can shrink a circuit. But, yet again, why move to 32NM when you just moved to 45NM? AMD.

If you look at it like this, Intel fanboys have alot to thank AMD for. Else, they wouldn't be an Intel fanboy they would just be another PC user. An 800 mhz single core 130NM (if that) PC user. You can argue that it might have been higher but not too much. Without anything to compare it to, how long would we have been stuck with the P4? How would we know that there was more to power than uber-clock speed with insane temps and power consumption?

You can say Intel is the chip for you and right now they are awesome chips (bought my first C2D about 3 months ago and love it) But, I will never forget my athlon which was cheaper and far better performing at a lower power level then the best Intel had to offer. I hope AMD gets back to that, because they cannot last much longer. Intel could do with another dunk in the tank. Maybe enough to make them go in the Red for a few quarters (they can survive it) then we would see just what the performance/price war would look like.


RE: Sweet!
By Mattz0r on 7/26/2007 2:30:54 PM , Rating: 2
I suppose you also have to define what doing "well" for Barcelona is going to be. If "well" means not losing more market share to Intel, and just staying afloat until all these new products come to market, then I'm pretty sure it'll do well. If, on the other hand, AMD is hoping that Barcelona is going to win them back all the market share they've lost in the past year, and that's their baseline "well", then I think yeah, it's a BIG if.


RE: Sweet!
By Tsuwamono on 7/26/2007 2:33:53 PM , Rating: 2
WTF does that do with anything? your still sucked in by the numbers. I'm an AMD fanboy running an E6600 at 3.4ghz. So what you run has nothing to do with what you prefer.. i PREFER amd because of several reasons but i just liked the Intels numbers for video encoding and stuff. If i only gamed i would definitely use AMD only


RE: Sweet!
By webdawg77 on 7/26/2007 2:43:26 PM , Rating: 2
quote:
WTF does that do with anything?


It implies that I did just choose to bash on the previous poster just because I like or dislike Intel (the presumption of liking in this case).

I go for the best price / performance ratio. When I built the PCs 2-3 years ago, AMD had that. If I was building one today, Intel would be the choice.

Do people have to explain things in great detail all the time for other to get it?

All I said was that Barcelona needs to do well to bring AMD back to a competitive level. Otherwise, it's going to be bad for both AMD and consumers alike.


RE: Sweet!
By zpdixon on 7/27/2007 4:13:48 AM , Rating: 4
Dailytech readers are generally enthusiast users, much more informed about technology than your average Joe Six-Pack. And I find surprising that most of them seem scared about AMD's upcoming K10 generation of processors not being able to compete against Intel's Core2 processors. I have a message to pass to everybody: K10 will assuredly regain the lead over the Core2 microarchitecture .

Given all the many tidbits of information I have grabbed from AMD's engineers over the past 12 months, I feel confident I can go as far as predicting that, at similar clock speed, K10 will be about 50% faster than Core2 on floating-point operations . In other words the quad-core Opteron x350 (2.0GHz, to be released in August 2007) will be, FPU-wise, about as fast as the highest-end quad-core Xeon X5365 available today (3.0 GHz). And of course HT and IMC will give the Opteron this extra lead to keep it a little bit ahead of the Xeon, especially in server workloads. A bit later, in Q4'07, AMD will ramp K10 to 2.4/2.5 GHz, which will give them a comfortable advantage over Intel's rumored "best" of 3.2/3.3 GHz (Intel would need 3.7/3.8 GHz to match AMD's performance).

Anybody familiar with today's CPU microarchitectures know how and why Intel has the lead, today, with Core2. Heck, I even remember explaining to people in 2005, when Intel's Netburst was being trashed by AMD's K8, that they absolutely had to adopt a true 3-way superscalar uarch if they wanted to remain competitive. And guess what ? It's exactly what they did with Core2.

So, of course, today AMD also know exactly what they need to do to remain competitive. A few hints to those who really don't know what I am talking about: AMD will improve the instr throughput / clock (solved by adding more gates per execution unit, especially SSE/FPU), reduce instr latencies of many integer/SSE/FPU instr from 3 to 1 clock , and maybe increase the associativity level of the L1 cache from 2-way to 8-way, etc.

Just wait and you'll see ;-)


RE: Sweet!
By crystal clear on 7/27/2007 6:55:08 AM , Rating: 3
quote:
Just wait and you'll see


Yes indeed we have been waiting & still waiting & will continue to wait.

All we saw is roadmaps & announcements !

People are getting pretty bored waiting !


RE: Sweet!
By Amiga500 on 7/27/2007 9:06:20 AM , Rating: 2
Yeah, thats something along the lines of what I've heard as well.

But, damn, I want to see it officially - I need a new workstation, and holding off is killing me!


RE: Sweet!
By Andrwken on 7/27/2007 11:15:41 PM , Rating: 2
and waiting and waiting and waiting. Benchmarks and some tangible evidence of how the k10 will perform would be great.

This article is a great PR move to keep the juices flowing while the current architecture stinks up the place with no information to speak of. It's supposedly only a few weeks from release, and noone really has a clue how good it's supposed to be. Yawn, this would be much more interesting if I knew more about this year's new product and not getting promises on the one 2-3 years out.


RE: Sweet!
By defter on 7/26/2007 2:43:05 PM , Rating: 2
What awesome stuff AMD has introduced recently (i.e. in the past 6 months)?


RE: Sweet!
By crystal clear on 7/27/2007 6:42:44 AM , Rating: 1
yes-some awesome announcements ! ha


RE: Sweet!
By RW on 7/27/2007 4:47:29 AM , Rating: 2
Now expect Intel to release an 8 core CPU in early 2008


RE: Sweet!
By Hawkido on 7/30/2007 4:55:32 PM , Rating: 2
I don't think their FSB can handle it...

Their Quad is a sandwitched Dual core with another dual core... sandwitching the same thing again will only give limited returns. As the CPU's will have limited knowledge (if any) of what the other core is doing. Intel will need to make a native Quad then sandwitch it or go native 8. but then the FSB will be depleted and suffer. so they need to redevelop the FSB (or move to something like hypertransport) to feed the processors.

I would like to hear what some of our more learned readers have to say about intel doing an 8 core and what they would need to change to compete with an AMD 8 core. Did they mention if the AMD was native or two sandwitched 4-core dice?

I think AMD will truely shine in the high core count server market around 2010. Intel will need to get all their stuff certified if they radically change their processor/MB setup before it will be accepted in the business world, because if you screw up there... Well you just screwed up everywhere.


AMD's Strong Point
By Goty on 7/26/2007 3:16:32 PM , Rating: 3
This is where AMD's design choices will really start to shine. Being built from the ground up to support more and more processor cores allows AMD to essentially just tag on more cores to the die at will, they've just got to wait for their process technology to catch up an mature to make it cost effective for the end users.




RE: AMD's Strong Point
By DeepBlue1975 on 7/26/2007 4:16:43 PM , Rating: 2
But the problem of native multicores lies in harder to get yields.
And with AMD not exactly having the upper hand on process technology, I wonder if it's the right moment for them to go full on native multiple cores.
I used to criticize Intel's "glue'em up first, nativitize them later" but now I understand it as being an enabler to a more rapid growing on core count while not risking yields and, of course, it also enables them to have a shorter time to market for these parts.


RE: AMD's Strong Point
By dgingeri on 7/26/2007 4:41:55 PM , Rating: 2
This approach yields faster 'to market' times, but gives diminishing returns while not reducing the costs. Their shared bus makes the Pentium and Core lines vulnerable to bandwidth hogging by a single application or process. The AMD approach is far superior, even in cases of a single memory buss for 2 or 4 cores, because it would scale much better going to multiple sockets.

When AMD went with 3 execution pipelines to increase efficiency per clock cycle, Intel ramped up the clock speed. When AMD introduced a single bus per chip, Intel increased the clockspeed on their shared bus. Intel has a history of simply using the brute force method instead of trying to find a better way to do things.

Then there is the little cheat that Intel tried to pull with Rambus. They knew it wouldn't perform better than SDRAM, but they had a deal with Rambus. They pushed it for their own financial gain and using their market muscle, not trying to push the marketplace forward to newer and better things. (Yes, I know corporations are supposed to concentrate on making money, but pushing forward the industry as a whole is a much more effective approach for long term returns than trying to rob the marketplace blind.) I still haven't forgiven them for that garbage.


RE: AMD's Strong Point
By Ringold on 7/26/2007 7:29:49 PM , Rating: 2
quote:
This approach yields faster 'to market' times ... The AMD approach is far superior


That's great and all, but hate to point out, as others have, that while AMD's practice may be technically superior they can't go to the local bank branch and exchange superior chips for cash. Only the marketplace allows that transaction to happen, and if their parts are in the labs while Intel's are out on the market... AMD can afford to take the moral high road to appease enthusiasts for only so long.


RE: AMD's Strong Point
By Goty on 7/27/2007 4:12:53 PM , Rating: 2
I agree with the point on yields, and that's why I mentioned the fact of needing the time to let the process mature.

I forget exactly what AMD calls it, but supposedly they have a system to automatically detect ways to improve the yields from wafer to wafer and they can automatically implement them without having to shut down part of the fab to retool or reprogram.

I'm not entirely sure of how it works and I'm sure that my explanation is wrong on a lot of levels, but I think this is the way that AMD was able to remain somewhat competitive with Intel in volume (when the demand for AMD's chips was still booming) even considering it's large disadvantage when it comes to numbers of fabs and the capabilities of each fab in particular.


Core Battles in Full Stride
By EndPCNoise on 7/26/2007 2:33:52 PM , Rating: 2
GHz wars are history. It seems Fusion did not end the core wars.




RE: Core Battles in Full Stride
By Warren21 on 7/26/2007 2:44:52 PM , Rating: 2
It's kind of hard for it to 'end the core wars' without being here...

(Reading between the lines; Fusion has been announced, but its nowhere near production yet, let alone released)


RE: Core Battles in Full Stride
By Amiga500 on 7/26/2007 5:40:04 PM , Rating: 3
The dark side clouds everything - begun... the core wars have.


Talk? Talk is cheap!
By ChipDude on 7/26/07, Rating: 0
RE: Talk? Talk is cheap!
By redpriest_ on 7/26/2007 4:37:29 PM , Rating: 2
Apparently you forgot all about Prescott and the promises and the glowing praises about 4 ghz and beyond.


RE: Talk? Talk is cheap!
By ChipDude on 7/26/2007 4:49:24 PM , Rating: 1
To reflect back on history is only usefuly if it is relevant in the context of the future.

Today, its what a month before release I find it interesting not one benchmark on barcelona. At this point if they can talk so much about something so far out they could show what the got in their hand. I don't deny INTEL didn't do a lot of talking during their Prescott missaventures. 4Ghz wasn't the problem with Prescott it was the power problem at 4GHZ. INTEL got over that and once they fixed it with Core2 they showed systems and even let outsiders play while being supervised. TOday INTEL has been more open then anything AMD has done. AMD is now just "talking" again. Its funny that today AMD is so close to release and nothing but more "talk."

If you had a postive bombshell wouldn't you unlease it to help your credibility. THey sure could use it at the moment after losing another 600million. Its not like INTEL can release Nehalem or Penrym tomorrow to respond. YOu are a month away show the goods! It ain't like your wife to be holding out till Honeymoon.


RE: Talk? Talk is cheap!
By Ringold on 7/26/2007 7:34:23 PM , Rating: 3
Hector Ruiz was on Closing Bell today and gave a distressingly horrible performance. Not a word of substance regarding AMD's business performance, but about a thousand hyperbolic words like "fantastic" thing and "great" that; he even said they were "very excited", if I recall, about the current corporate bond market. I'm sorry -- who the hell gets excited about bonds, ever? :P

He did say insiders were very enthusiastic about Barcelona, but c'mon, that could mean anything.


I wonder if the will stack two Barcelona's?
By DallasTexas on 7/26/2007 3:02:24 PM , Rating: 2
Hmm. I wonder if AMD will stack 2 quads. It seems their "fusion" product line includes stacking graphics and CPU.

So the "Native this and that" baloney laughed at by the industry, is most likely comming to an end. Looks like AMD's "next invention" is stacking die together with some clever name, of course.




By ceefka on 7/31/2007 6:47:29 AM , Rating: 2
quote:
stacking die together with some clever name, of course.


It's called gluesion.


Good ot know AMD is still all talk
By bull2760 on 7/26/2007 5:33:59 PM , Rating: 3
Why bother bringing up future plans about an 8 core processor, they can't even get barcelona out without any problems. Hector Ruiz needs to go, heads should be rolling for all the delays in the product developement, and yet he does nothing. Someone needs to be held accountable for all the bad moves AMD has made, and you always start at the top.




By DeepBlue1975 on 7/26/2007 3:02:07 PM , Rating: 2
I think they should concentrate on deliverying barcelona and getting it right before leaking out roadmaps that talk about wholly new architectures.
It'll be very embarrassing and image damaging for them if projects like this end up being dead before they are born, like the k9 or whatever its code name was.

Their wheel is not spinning now and they need it to start moving again before talking about running at 200mph.

PS: I'm no Intel fanboy, as a matter of fact, I'm still using an athlon 64 3200+ with 1gb of RAM since 2005 and am only recently being tempted to change to an Intel Q6600, but am not fully decided as of yet.




By DeepBlue1975 on 7/26/2007 3:02:25 PM , Rating: 2
I think they should concentrate on deliverying barcelona and getting it right before leaking out roadmaps that talk about wholly new architectures.
It'll be very embarrassing and image damaging for them if projects like this end up being dead before they are born, like the k9 or whatever its code name was.

Their wheel is not spinning now and they need it to start moving again before talking about running at 200mph.

PS: I'm no Intel fanboy, as a matter of fact, I'm still using an athlon 64 3200+ with 1gb of RAM since 2005 and am only recently being tempted to change to an Intel Q6600, but am not fully decided as of yet.




What was Mimzy?
By TimberJon on 7/31/2007 4:18:53 PM , Rating: 2
I know what a nanometer is, but what wasnt explained was how far did the nanotech go with the circuitry buried in that rabbit from the Mimzy movie. They took a slice off the bear, and managed to zoom in with I believe an electron microscope several levels of zoom.

I don't even know if that was completely sci fi and just meant to wow the audience, or if it is physically possible to manufacture down to that size.




AMD shows off 3.0GHz Barcelona
By End Of Times on 7/26/2007 4:06:32 PM , Rating: 1
"This is about the Internet.  Everything on the Internet is encrypted. This is not a BlackBerry-only issue. If they can't deal with the Internet, they should shut it off." -- RIM co-CEO Michael Lazaridis

Related Articles
AMD Announces G3MX Technology
July 26, 2007, 8:00 AM
AMD's 45nm Opterons Scheduled for 2008
February 5, 2007, 11:50 AM
PCIe 2.0 Ratified
January 16, 2007, 5:54 AM













botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki