Monday, July 24, 2006

AMD's denying move

As I proposed in this article titled "AMD: time for the kill", AMD must deny Intel the oppurtunity to unload its Netburst inventory at a good price. Specifically, I proposed that AMD should have low priced dual core to stop the Pentium D dumping, and a low priced Sempron to deny Celeron the market. Now, AMD has published its new pricing list. The AM2 X2 3800+ is priced at $152 and the AM2 Sempron 2800+ is priced at $47. With this pricing, one can safely say, Intel's $8.66 billion Netburst inventory is trashed overnite. And, Intel's FABs are producing more and more unwanted Pentium D chips.

84 Comments:

Blogger Mad Mod Mike said...

I think for the consumers that know about AMD and Intel specifically this will benefit the most, but I fear the people who do not know will buy whatever is known to them; sadly this is Intel's Pentium 4.

The real gold strike will be business's such as HP and Dell that could hopefully convert from selling worthless Pentium 4's and start giving their customers worthy chips like the Athlon 64.

5:37 PM, July 24, 2006  
Anonymous Anonymous said...

If only the turion x2s were priced more in line with the new athlon x2 pricing. Performance-wise, the turion x2s are beat out even by yonah. (yeah I know, yonah is only 32-bit, yadda yadda yadda...)

5:38 PM, July 24, 2006  
Anonymous Anonymous said...

I, me, mine, my. Your favorite words. You are a legend in your own mind.

You won't publish this, but you can at least read it.

5:41 PM, July 24, 2006  
Anonymous Anonymous said...

Intel comes to a gun fight with a knife. I was among those in disbelief upon first hearing about of the merger. It seems like everyday Intel has to eat more shit for lunch and they thought they were having SuperPi.

What so intriguing about the chess match, it seems Hector has anticipated every move Paul makes. From the onset the analyst were giving the win to Intel by virtue of its size and glorious production history. I always thought “size matters”.

6:13 PM, July 24, 2006  
Anonymous Anonymous said...

DOC,
SO YOU ARE NOT DENYING YOU WERE FIRED FROM INTEL?

6:53 PM, July 24, 2006  
Anonymous enumae said...

Could you tell us what percentage are Pentium D.

And how do you know they arn't producing Conroe, Merom and Woodcrest?

"sadly this is Intel's Pentium 4."

I agree with MMM, people don't know about AMD. Anytime I watch sports I see an Intel comercial, AMD has poor advertising.

6:55 PM, July 24, 2006  
Anonymous Richard Evans said...

I have to agree with mad mod mike, that the average consumer doesn't even know AMD exists, or if they do, they don't realise they make an alternative CPU to "Intel Inside".

I've fixed enough systems using P4 2.4's for co-workers to realise that NONE of them know about AMD. I've done everything I can to educate everyone I work with, but there's still a general ignorance by the non-computer-savvy public. (Although my current boss surprised me, by being completely computer ignorant, but knowing about "these new 64 computers" a couple of years ago and asking me about them.)

I digress--AMD, even with the price cuts, (UNTIL Dell starts selling them regularly) will suffer from "what's that?" status with the general public. And let's face it, that's the market the obsolete P4's are going to be aimed at.

7:04 PM, July 24, 2006  
Blogger Sharikou, Ph. D said...

I have to agree with mad mod mike, that the average consumer doesn't even know AMD exists, or if they do, they don't realise they make an alternative CPU to "Intel Inside".

But 80% of the people bought AMD PCes in stores -- that's a fact. Those who don't know AMD must be the 20%.

7:06 PM, July 24, 2006  
Anonymous Richard Evans said...

"But 80% of the people bought AMD PCes in stores -- that's a fact"
Do you mean from Bestbuy and its ilk or from boutique vendors such as VoodooPC or Falcon Northwest? I could believe it from the boutique vendors, the computer savvy buy from them, but bestbuy or here in Canada, future shop?

I'm a big AMD fanboy, and my next sys will be am2, but while I realise that they've made up some ground, I haven't seen statistics that indicate they've caught up quite *that* much. Could you expand with a link to show that the consumer desktop space has been *that* dominated by AMD at any point? I'd appreciate seeing those stats, it would make me cheer!

7:17 PM, July 24, 2006  
Anonymous Anonymous said...

Netburst inventory will still clear, because the products are still well priced For instance the 945D is faster than the X2 3800+ and is priced accordingly. About $10 more than the X2 3800+. Therefore, the 945D is still a very attractive option and has perfectly competitive price/performance. This is the same for the 805D for example.

7:27 PM, July 24, 2006  
Blogger Sharikou, Ph. D said...

Do you mean from Bestbuy and its ilk or from boutique vendors such as VoodooPC or Falcon Northwest?

Google "AMD retail PC market share current analysis". It's well known that AMD dominates the US retail market.

7:40 PM, July 24, 2006  
Blogger Sharikou, Ph. D said...

Netburst inventory will still clear, because the products are still well priced For instance the 945D is faster than the X2 3800+

X2 3800+ frags Pentium XE 965, an X2 3600+ will frag a Pentium D 945.

7:42 PM, July 24, 2006  
Anonymous Anonymous said...

yeah right...

AMD needs some revenue. They are already selling their 90nm near breakeven. Go any lower the losses will mount and how will they finance their 65nm 300mm factory, there 45nm development.

INTEL has already made their 65nm capital spends and the bulk of their 45nm too.

AMD execs are stupid, but thankfully not as stupid as you.

Don't confuse fanboys dream of a cheap CPU with a broken business strategy for AMD
The Doctor

8:09 PM, July 24, 2006  
Anonymous Richard Evans said...

Google "AMD retail PC market share current analysis". It's well known that AMD dominates the US retail market.

I did your google, and the closest thing I found was an x-bit labs report from another source that indicated that in the low-end segment, up to $750, AMD dominated (REALLY dominated, up to almost 90%!), but higher than that, Intel led in the next segment (not by much) and in the ultra-high end, Intel dominated with over 90%.

A further question: are online or "over the phone" orders included in this report, or is this pure retail? Do Dell sales get involved in these numbers? Or because they do a lot of their sales over the internet or over the phone skew the results?

8:10 PM, July 24, 2006  
Anonymous Richard Evans said...

Quick add-on-- the best results for overall consumer retail by AMD I've seen have been 53%. Which is fantastic! This was following your google suggestion.

8:14 PM, July 24, 2006  
Anonymous Anonymous said...

Sharikou,

You and other do not understand this move correctly.. First, timing is everything.. AMD did this move at this time (and not sooner) to steal the show from the intro of CONROE..

Second, this is AMD's way of throwing the towel and saying "we want to be a platform company" just like the other guys..

Think about it.. Look at AMD's history.. They introduce a superior product.. They win for a little while.. Then comes Intel and beat them at their own game.. What does AMD do? Change the rules of the game.. WHen they could not compete on GHz, they changed their scheme of numbering aka ****XP.. WHen they lost the performance crown, they are adopting the platform vision AKS Torrenza.. The way I see it is one way: AMD is dropping our of the race and ATI will definitely be a tough job for them..

A good company may not be easy to integrate into the core business.. When you run out of ideas, buy a company! remember HP Compaq DEC ..

9:26 PM, July 24, 2006  
Anonymous george said...

AMD can survive losses longer than intel, because of there cost structure, and AMD's chips are ~= to Intels Chips, die size wize. and they cant sell CONrow to make up for the ground they lost in channal clearing. I ben in office max, in the desktop market i saw about 10 AMD pc's to 1 Intel PC, so the only market i worry about is Corpret Desktop.

9:36 PM, July 24, 2006  
Blogger Sharikou, Ph. D said...

First, timing is everything.. AMD did this move at this time (and not sooner) to steal the show from the intro of CONROE..

Conroe is no threat. AMD is not concerned with Conroe at all. In fact, Conroe did something good for AMD. It's Conroe which pushed DELL to AMD sooner. Also, Conroe will effectively kill Intel -- you have to keep in mind that Intel's 90% production is Netburst, even today.

Script kiddies don't understand the economics.

9:41 PM, July 24, 2006  
Anonymous Anonymous said...

Note to the “The Doctor”

I love your posts. You always come to the blog with you panties in wad. Relax.

Intel Health Plan Administrator

9:56 PM, July 24, 2006  
Blogger Michael said...

I'm also afraid that the vast majority of PC buyers don't even know anything about what they are buying. They don't care weather it is AMD or Intel because they don't know that they both exist.

CyberSurge

9:57 PM, July 24, 2006  
Anonymous Anonymous said...

"A further question: are online or "over the phone" orders included in this report, or is this pure retail? Do Dell sales get involved in these numbers? Or because they do a lot of their sales over the internet or over the phone skew the results?"

Most of the "official" retail market segment #'s do not include online sales (including Dell). Although I cannnot speak to this specific report.

10:08 PM, July 24, 2006  
Anonymous enumae said...

"I always thought “size matters”."

Any woman who tells you diferently is lying...lol

10:09 PM, July 24, 2006  
Anonymous Anonymous said...

"X2 3800+ frags Pentium XE 965"

You keep providing that same link and it does not support your point of view. If you look at ALL of the benchmarks in that link it is clear 965 outperforms the 3800, you have just cherry picked the one page out of I think 8-9 pages of benchmarks in that same review that shows 3800 over the 965.

Are you just hoping that the average reader is just too lazy to look at any of the other benchmarks in the article you provided or are you just so lazy that you didn't read the rest of the article?

10:14 PM, July 24, 2006  
Anonymous Anonymous said...

"Quick add-on-- the best results for overall consumer retail by AMD I've seen have been 53%. Which is fantastic! This was following your google suggestion."

Was this US market only? I believe US market is now <40% of overall CPU sales, when you factor in the % of market that is retail the market gets even smaller on % basis.

10:18 PM, July 24, 2006  
Anonymous Anonymous said...

Sharikou: "you have to keep in mind that Intel's 90% production is Netburst, even today."

How many times are you going to keep spreading this misinformation?

90% of Intel's current production is not Netburst, as a large chunk of Intel's production is also Core (current mobile product? Yonah), and the roughly 10% Core 2 you have been estimating. Do you not agree with this? I believe this is the 3rd or 4th time I have tried to correct it (in fact you even updated one of your blogs after I mentioned this once!)

Do you believe if you keep printing that same incorrect statement it might become true? Please try to show a modicum of integrity.

10:24 PM, July 24, 2006  
Anonymous Richard Evans said...

Was this US market only? I believe US market is now <40% of overall CPU sales, when you factor in the % of market that is retail the market gets even smaller on % basis.

It was a reference not to overall retail, but desktop retail. And as I said in my previous post, I don't know that those numbers included internet or direct from the manufacturer, namely Dell...(often these statistic-takers only include people going to a physical store and buying) Intel/Dell would obviously have a large impact if direct-mail wasn't included in the statistics.

But still, considering where AMD was three years ago, it has been an excellent improvement overall, and Dell selling AMD in the consumer space *will* make the situation far more competetive (even with Intel trying to offload their crappy P4's as fast as possible), which can only be good for the consumer.
Let's face it, the average consumer might not have known who AMD was or what they did, but if nothing else, they're starting to know. My parents, for example, bought a computer without asking my opinion, and when they got it, they were disappointed with its performance (My father was the buyer, and he thought a 52X CD-RW must be better than an 8X Dual-layer DVD-RW--52X is faster than 8X!!). Since I'm 600 km away from them, they didn't think to ask at the time, but now they won't stop bothering me with questions about how to upgrade it. I'm pretty sure their next system will be AMD, but they're a perfect example of the uneducated consumer purchasing Intel out of complete ignorance, using brand marketing as their guide. AMD had cheaper and faster solutions out at the time they bought their system (Celeron 2 ghz).

Wow, that post meandered--sorry!

2:59 AM, July 25, 2006  
Anonymous Anonymous said...

ROTFL.

"Conroe is no threat. AMD is not concerned with Conroe at all."

Compare to:

"No I am not scared, and neither should you be!"
"There are no American infidels in Baghdad. Never!"

Similar, eh?

The fact is that AMD can no longer sell their overpriced pre-war K7 chips with integrated memory controller they prefer to call K8.

3:48 AM, July 25, 2006  
Anonymous Anonymous said...

Well, I can see that the EE 965 is fragging the X2...
And where do you get the "8.66 billion inventory" from?

4:23 AM, July 25, 2006  
Anonymous vandexter@libero.it said...

Conroe may be faster in MP3 encoding and zipping files, but the -very important because enthusiast- gaming performance is very close to the FX62. Maybe Intel can put some higher clocked Conroes onto the market but it isn't really that superior for beeing a whole new so called futureproof architecture. K8L is coming out soon.
Check out this link:

http://enthusiast.hardocp.com/article.html?art=MTEwOCwxLCxoZW50aHVzaWFzdA==

Note that in this test has been used DDR2 800 Corsair memory. The guys shoul repeat the test using higher clocked memory such as 1066 and we will see an X6800 and the FX62 beeing equal in performance, with the FX beeing a 2,8 and the X6800 a 2,93 CPU.

The corporate (volume-) market obviously is very important, but you really don't need high performance for office applications (under XP) and nobody cares or even knows about which CPU is working inside. The only thing that matters is the price of the system. Just ask your boss what he is thinking about this. If AMD/ATI can provide a stable and cheap office platform they will gain a lot.

The gaming market is the most important for performance CPUs and if Conroe is all Intel can throw in than I see a further dominance for AMD the coming years. Specially now togheter with ATI they can for shure develop astonishing products for the enthusiast market.

The notebook market is the fastest growing market and as already said ATI provides the best platforms (chipset & grfx).

Don't wanna talk about the server segment because there has already been said nearly everything ;-)

Not to forget VISTA! Microsoft seemed to be really enthusiastic about the merger, for shure they know something more and are obviously very interested in a very fast spreading of the new OS on new fast platforms.

4:33 AM, July 25, 2006  
Anonymous Anonymous said...

mr. shari cow, i've just seen a review at xbitlabs that 64bit performance of core2 micro-arch almost always is superior to athlon/opteron. what say now? btw, i predict a doom of both amd/ati in 5Q!!

7:11 AM, July 25, 2006  
Anonymous Anonymous said...

With regards to retail sales, I would like to make a clarification: retail sales include only sales made in retail stores (Best Buy, Circuit City, etc). This amounts for an incredibly small portion of the PC market. Most PCs are sold by vendors such as Dell, HP, etc.

And yes, Intel all but left the retail segment of the PC industry, and don't expect them to return any time soon. The knowingly send very limited quantities of their chips to the retail segment in order to meet the demand in other segments.

8:12 AM, July 25, 2006  
Anonymous Anonymous said...

re:A good company may not be easy to integrate into the core business.. When you run out of ideas, buy a company! remember HP Compaq DEC ..


This deal has the potential of expanding businesses in two fast growth areas: Digital TV, handhelds where ATI has a good presence.
See the AMD interview at Digitimes.

Buying ATI might be a way to put some of the ideas in motion.

How about an upgraded version of the OLPC laptops for US consumption in the future with joined AMD/ATI technologies?

HP's buying of Compaq turned out quite well for HP. When HP finally implemented all of Carly F's plans like major layoff and expense cuts, it narrowed Dell's cost advantage.

9:18 AM, July 25, 2006  
Anonymous Anonymous said...

Sharikou,
The techreport link you provided does not show x2 3800 beating 965. The 3800 does win a far number of benchmarks, but by no means a majority. Looking at ALL of the pages of that review, I'd say the 3800 and 965 are pretty much on par, excelling at different things. Please stop claiming that the 3800 "frags" the 965 as though it were a fact. Not even your own evidence supports that claim.

10:37 AM, July 25, 2006  
Anonymous Anonymous said...

Intel's 90% production is Netburst

A URL pulled up from google (an article dated June 28, 2006) says that, in the desktop space "about 10%" of the processors shipped by Intel in Q3 will be Conroes; in Q4, about 30% will be Conroes.

The same article says that in the notebook space Merom will account for 10% in Q3, and more than 50% by end Q4.

In the server space, Intel expects Woodcrest to make up 50% in Q3 and more in Q4 ... but you know, when companies are becoming less fond of your previous product, it's not a hard business decision to produce less of the crappy product that people were buying less of anyway.

Overall, who knows what the Core Microarchitecture share of total production is vs. Netburst.

11:37 AM, July 25, 2006  
Anonymous Edward said...

It doesn't seem like a desparate move of AMD to buy ATI; otherwise it won't spend that much cache on the deal. It seems AMD really isn't afraid of Core 2 Duo at all. That's somehow puzzling, because Core 2 Duo will occupy the spotlight till the end of the year, which will include also the back-to-school and Christmas buying seasons. 4x4 is granted by AMD executives to be enthusiast-centric only. It simply doesn't ring for most consumers who are cost-sensitive.

AMD probably knows something that we don't. Either Intel's production won't ramp up as it claimed, or AMD still has ways to improve its K8 core performance (Rev.G?) before end of the year. Or maybe 2/4 socket server market is growing well and AMD relies less and less on high-end PC desktop sales.

Another big reason could be that AMD is eager to expand its R&D capability to compete with Intel. It might be more efficient to buy the whole team (from ATI) than hire engineers one-by-one. This could be important because Intel is really attacking AMD at all fronts: on media-center PCs, on memory architectures, on CPU core design, on multi-core, and on mobile and server specific architecture. AMD with its limited resource & revenue has been having a hard time to cope with these. OTOH, it's known by many that 1) ATI has a number of good engineering teams with more potentials than they appear, 2) ATI doesn't have brands as strong as nVidia (read: nVidia's marketing is too much stronger). Looking from these perspective the merger with ATI (instead of nVidia) makes much more sense.

11:48 AM, July 25, 2006  
Anonymous Anonymous said...

"Sharikou, Ph. D said...

Conroe is no threat. AMD is not concerned with Conroe at all. In fact, Conroe did something good for AMD. It's Conroe which pushed DELL to AMD sooner. Also, Conroe will effectively kill Intel -- "

Well I've been a supporter of AMD for a long, long time... however, after the recent benchmarks that have come out I would say Conroe is indeed a threat to AMD and a big one at that. I will definitely be giving it a run I think as my next cpu.

12:14 PM, July 25, 2006  
Blogger TheKhalif said...

More importantly AMD is about to release the X2 3600+ for under $150. It is 2GHz with 2x256 L2.

This will be the best machine out for awhile in terms of price. Supposedly Mitac is assembling systems now for Dell(?) and I kknwo they will be including these even if only for HP.


Intel is going to have a helluva time getting rid of HeatBurst no matter what AMD does.

They shot themselves in the foot with Core 2. it's too good to be priced the way it is. It totally devalued the whole desktop market which is another reason for this purchase.

Turion X2 can replace Core Duo in a lot of systems and with an ATi mobiel chipset.....

1:11 PM, July 25, 2006  
Anonymous Anonymous said...

All I can say is that,it is the beginning of sleepless nights for Intels execs !!

1:12 PM, July 25, 2006  
Blogger TheKhalif said...

I digress--AMD, even with the price cuts, (UNTIL Dell starts selling them regularly) will suffer from "what's that?" status with the general public. And let's face it, that's the market the obsolete P4's are going to be aimed at.


That's not true. Peopel just don't care what CPU is in their PC. tehy only care how much it costs. That's why Sempron is the most popular retail system.

Even people who recognize the Intel chime probably STILL don't know what Centrino is.


AMD has some really cool ads for businesses in downtown NYC. They just don't push the desktop because Opteron is worth more per sale and willalso give shops more confidence in X2.

Build teh server and they will come.

I can guarantee that AMD will relase a glueless chipset right before K8L.

That's what they want. Broadcom does most of their 4P chipsets for new machines.

They can probably still use 8000/8131, but it's getting old and they need a chipset built for HT3.

1:16 PM, July 25, 2006  
Blogger TheKhalif said...

Note that in this test has been used DDR2 800 Corsair memory. The guys shoul repeat the test using higher clocked memory such as 1066 and we will see an X6800 and the FX62 beeing equal in performance, with the FX beeing a 2,8 and the X6800 a 2,93 CPU.


anand has done this test. he clocked the FX bus to 266 to run the RAM at 1066. Core 2 still wins but it is closer.

Both chips are at 2.9 and RAM is at 1066. I din't expect any different. Core 2 IS faster. it doesn't leave it in the dust the way FX 60 did to 955/965, but it is a win.

Looking at the specs of K8L, I can say enjoy it while you can. That architecture is 20-30% "better" than Core 2. an dthat's not taking into account the shared L3.

1:26 PM, July 25, 2006  
Anonymous Anonymous said...

Well I've been a supporter of AMD for a long, long time... however, after the recent benchmarks that have come out I would say Conroe is indeed a threat to AMD and a big one at that. I will definitely be giving it a run I think as my next cpu.

my oh my. another paid intel pumper! *wah!*

1:44 PM, July 25, 2006  
Anonymous Anonymous said...

Well I've been a supporter of AMD for a long, long time... however, after the recent benchmarks that have come out I would say Conroe is indeed a threat to AMD and a big one at that. I will definitely be giving it a run I think as my next cpu.

Its availability and similarly price which will be most important for conroe. After all - what does it matter if a CPU nobody can get beats you? The fact that Intel isn't producing only CORE chips by now is scary.

2:20 PM, July 25, 2006  
Anonymous The Power of Now said...

AMD is on the run, taking hit after hit from Intel.

AMD's buy of ATI was a DEFENSIVE move.

AMD's cut-rate-box new CPU pricing was a DEFENSIVE move.

AMD's two-slow-chips-are-better-than-one-fast-chip is another DEFENSIVE move.

Sure there is the pie in the sky K8L, but that might as well be:

KL8 because it is certainly LATE.

AMD should have had new Opteron architecture 2-3 years ago. It will have been 7 years since the original Opteron design. Pitiful.

AMD has nothing interesting to offer the market TODAY. Just a bunch of science talk about Torrenza, 4x4, KL8, etc. None of which will be meaningful until 2008 sometime.

Meanwhile if Intel keeps up with the body blows, AMD is going down for the count. Because shipping the world's fastest chips, TODAY, is what counts.

4:54 PM, July 25, 2006  
Anonymous Anonymous said...

Why AMD got no worries.

Benchmarks, they don't matter except to a few enthusiasts and they don't matter

TDP, the matter until we lose them then they don't matter either.

Profits, they don't matter, lets just drop our prices 50% and cut our revenue by a similar amount. Damm our 300mm factory costs and 45nm development cost don't matter. We'll just borrow some more money from Morgan STanley

K8L will save all, like we said benchmarks, power, profits don't matter. We'll just borrow another 5 billion in 2007 to buy nVIDIA when things don't work out.

Don't worry be happy

7:17 PM, July 25, 2006  
Anonymous Anonymous said...

AMD priced to kill. Where are all those Intel fanboys that were whining about price? Some said the merger was in your dreams, where are you? Most insisted Woody wasn’t a paper launch, where’s Woody and where are you? Words mean things.

7:42 PM, July 25, 2006  
Anonymous enumae said...

"After all - what does it matter if a CPU nobody can get beats you?"

Two more days and well know if your comment is justified.

"The fact that Intel isn't producing only CORE chips by now is scary."

How do you know, really how do you know that they arn't, I read an article saying Conroe could be pushed hard, up to about 50% for Q1 07. [Link]

I can't even remember a TV commercial from AMD, this is terrible...

Does anyone think that if/when Dell sells AMD we will see TV commercials without the Intel theme, and maybe an AMD theme instead?

8:26 PM, July 25, 2006  
Blogger SmartM0F0 said...

What Intel has essential done (even to Intel fanboys) is say that their P4 and netburst is complete crap compared to Athlons!!

Intel execs were under pressure from Wall Street to do something, and from Intel's history from as far as I can remember... they've resorted to paper launches just to claim they have the lead on something not out yet.

Just remember, starting 2nd half of this year, AMD will start having more capacity from Fab36, increase wafer starts at chartered, and this translates to more revenue even with a price war with Intel.

I think it is Intel that is in real trouble with its outdated P4 and netburst inventory. They have ONE QUARTER'S worth of inventory!! $4.3 billion worth!!

Understand this people:
- Comfortable (or lean) inventory is considered to be ~3-4 weeks in this industry.

- When your inventory hits 45-60 days, you should really start sweating no matter if you're the #1 company in the world or not.

- 60-80 days is considered RED FLAG territory.

- Intel, at the moment, has over 90+ days... and they got reamed a 2nd A-hole by analyst at their earnings asking if they are,... err... "comfortable" with going against history.

With the ATI announcement, and Conroe not supporting SLI at the moment... Intel will be selling Conroe/AMD Crossfires?? huh?

I'll wait for the day I see an AMD graphics sticker in front with the Intel sticker. :)

8:45 PM, July 25, 2006  
Anonymous Anonymous said...

vandexter@libero.it said...
Conroe may be faster in MP3 encoding and zipping files, but the -very important because enthusiast- gaming performance is very close to the FX62. Maybe Intel can put some higher clocked Conroes onto the market but it isn't really that superior for beeing a whole new so called futureproof architecture. K8L is coming out soon.

10:45 PM, July 25, 2006  
Blogger duploxxx said...

the power of now... all your statements are just crap....

The only thing that intel has are $ in old netburst and a paperlaunch core architecture... there are only "ghost" core archhitecture cpu's for sale called ES.

Intel will take the crown again after 3years of crappy chips. the desktop market is a low profit line all the way, always has been. budget wise the k8 is still a very attractive cpu to buy against the e6300 - e6600 chips. the server will be faster because they have a 3.0 and amd not at the moment. a 2.6 is as fast as a 2.66 woodcrest. power is the same price is the same.

by the end of this year 65nm will hit intel like a real hammer they don't need k8l for that. powerconsumption will be less then the core architecture (it is already less for all the chips except the highest speed with the ee version) on all of the chips and the die size is half of a conroe meaning that dropping the price again with 40-50% will still result in equal profit. fab 36 can produce enough waffers that fab 30 can convert and chatered can change to 65nm also.

intel cant drop prices enough until they start 45nm. and thats the only decent design they have for now... they can't stretch the design anymore.

1:13 AM, July 26, 2006  
Blogger Azary Omega said...

Conroe this , conroe that. Once those suckas are out im catching me one and roasting it in 775 socket. No, i wont keep it in doors for long. Just long enough to torture it in few test's and then - i'll sell it. Yup i'll sell it. To some intel loser who thinks that games in 800x600 are fun and incapable of counting pass the number 63....

-forever yours, Azary Omega

1:22 AM, July 26, 2006  
Anonymous Edward said...

"The fact is that AMD can no longer sell their overpriced pre-war K7 chips with integrated memory controller they prefer to call K8."

This is an extremely dumb statement because last time I checked NONE of any retail or online vendors sold a single system with Conroe. There are probably at most hundreds of Conroe retail box sold, but even double that, its marketshare in processors is effectively zero.

Unless you believe that none of those retail and online vendors made a dime from AMD systems over the past month, I see no merit in your claim that Conroe has the ability to stifle K8's sales.

All I can say is that YOUR claim of Conroe beating K8 sounds precisely like the claim of the Iraqi force beating the American invasion...

1:52 AM, July 26, 2006  
Anonymous vandexter@libero.it said...

I really don't read much into the standard benchmarks, just look at ATI vs. Nvidia and 3DMark. Nearly every time a new driver is released there is a gain of some points in that f**in 3DMark score. Obviously conroe has advantage due to the 4MB of Cache considering the benchmark programs don't require that much memory. In real world applications as gaming at high resolutions the story looks a bit different, I'm not talking about that stupid benchmarks. In the FEAR benchmark for example I get 69 FPS, something that is not true for the real gaming performance, same thing for Doom, Quake and all others. MP3 decoders and Winzip get some huge advantages due to Concroes 4MB cache, but I'm shure nobody ever uses Winzip all day for packing big files so this is a real stupid test. If I compress some files bigger than 50MB i don't look even at the time it needs. No matter if I wait a minute or a minute and ten seconds, that's ridiculous. And don't think there are some many people out there that buy a fast expensive system just to do some MP3 converting!
Fact is that in gaming AMD and Intel (now) are very close. Mainly the big work is still done by the Graphics adapter and not the CPU!
Most games have the SAME FPS on 10 different CPUs but using the same graphix! So even if you have only a Sempron 3200+ but a X1900 or 7900GTX inside games run really fast.
AMD/ATI are planning to introduce the first one-chip Computer by 2008, just read the statements of the conference. Intel tried to do this years ago with the Timna but failed.

2:44 AM, July 26, 2006  
Anonymous Anonymous said...

How do you know, really how do you know that they arn't, I read an article saying Conroe could be pushed hard, up to about 50% for Q1 07.

From your link:

"The new Intel roadmaps indicate that crossover (more new chips start being made than old chips) will occur sometime in the fourth quarter."

That means all the way until the fourth quarter (apparently), they will be producing more Netburst than Core.

Instead they should simply focus 100% on converting everything to Core right now. Even though their production will drop, they have plenty of inventory of Netburst to sell in the mean time. If they keep on this slow (relatively speaking) transition, they won't get rid of that gigantic inventory by the time Core is here in a much bigger way, and nobody bothers with the old stuff. Their total revenues won't really drop in the short term, but since they will be producing more Core chips sooner, in the mid term it will boost revenue.

4:59 AM, July 26, 2006  
Anonymous Anonymous said...

http://news.zdnet.com/2100-9584_22-6092226.html

Intel has started regaining market share-even before Conroe launches.

7:16 AM, July 26, 2006  
Anonymous Anonymous said...

Rahul Sood, president and CTO of VoodooPC said the AMD 4X4 will play two games at the same time. This was previously done with two separate PC’s. This is serious computing power.

http://www.crn.com/sections/breakingnews/dailyarchives.jhtml?articleId=191200462

7:31 AM, July 26, 2006  
Anonymous enumae said...

Looks to be only FX processors for 4x4, so much for it being cheap.

[Link]

8:26 AM, July 26, 2006  
Anonymous Anonymous said...

“AMD's buy of ATI was a DEFENSIVE move.
AMD's cut-rate-box new CPU pricing was a DEFENSIVE move.
AMD's two-slow-chips-are-better-than-one-fast-chip is another DEFENSIVE move.”

What’s wrong with defensive moves? It’s part of business. : )

Intel’s release of Yonah before 64-bit extensions could be made to work with it: Defensive.

EMT64, a copy of AMD’s 64bit instruction set, which often even translates to performance decreases: Defensive.

“It will have been 7 years since the original Opteron design. Pitiful.”

Finally squeezing out an architecture with only a 10-15% performance increase over an architecture that has been around for, as you said, 7 years. After years of selling 90-65 nm toaster ovens: Defensive.

Guess what? Pentium 4 architecture has been scraped after all that time, and the original Opteron design is still viable.

Pumping billions of dollars into the Itanium, doomed to failure, and still getting stuck without direct connect architecture on the higher scale server chips; Gee! That sounds like a great idea!!

“Sure there is the pie in the sky K8L, but that might as well be:
KL8 because it is certainly LATE.”

K8L is late? Why push your products out the door way too fast *cough* Intel *cough* when you don’t even have to in order to stay extremely competitive?

Intel better pray that it can keep from losing too much money on its Pentium 4 stock. Conroe and K8 outshine Pentium 4 incredibly, and Intel better put on some sunscreen.

9:20 AM, July 26, 2006  
Anonymous Anonymous said...

TheKhalif said;
"That architecture is 20-30% "better" than Core 2. an dthat's not taking into account the shared L3."

What makes you say that?

10:20 AM, July 26, 2006  
Anonymous Anonymous said...

Enumae. you're comparing a benchmark website information against intel's official information?

dude, even on intel's website, theres a flash animation wich shows perfectly and clear they will only supply 20%.. yes dear.. you read well this time. ONLY 20% OF CORE 2 DUO BY THE END OF 2006 ( and they expect it to increase it to 40% by the second quarter of 2007 )
go browse intel's website then come with this kind of crap

10:44 AM, July 26, 2006  
Anonymous Anonymous said...

AMD has nothing interesting to offer the market TODAY. Just a bunch of science talk about Torrenza, 4x4, KL8, etc. None of which will be meaningful until 2008 sometime.

And 2008 is when the pain really gets there for AMD. Nehalem. For AMD, that's spelled "D O O M".

People think AMD is feeling the pain now, what will they do when every single Intel technology is better than every single AMD tech? Right now, AMD has HT which is good stuff, and IMC but CSI is better and Nehalem will have IMC and CSI.

11:21 AM, July 26, 2006  
Anonymous enumae said...

"go browse intel's website then come with this kind of crap"

I was stating what I read, and it would seem to make sense due to the fact that they are not clearing there inventory of P4/Netburst.

12:25 PM, July 26, 2006  
Blogger TheKhalif said...

AMD should have had new Opteron architecture 2-3 years ago. It will have been 7 years since the original Opteron design. Pitiful.


The Opteron was announced in 2002 and released in Oct2003, quickly followed by true dual core in 2005.


AMD can pull a Core 2 fast one with K8L and get at least a 70% boost over current Opterons.


I am betting we'll see AMDs first new chipset in awhile for glueless 16Way+.

1:08 PM, July 26, 2006  
Blogger TheKhalif said...

You keep providing that same link and it does not support your point of view. If you look at ALL of the benchmarks in that link it is clear 965 outperforms the 3800, you have just cherry picked the one page out of I think 8-9 pages of benchmarks in that same review that shows 3800 over the 965.

Are you just hoping that the average reader is just too lazy to look at any of the other benchmarks in the article you provided or are you just so lazy that you didn't read the rest of the article?



SO it's okay for Core 2 at $300 to be faster AT ANYTHING than a $800 and it's ownage, but if AMD does it, it doesn't matter because it's not every test?

And remember the X2 3800+ is now $150.

1:11 PM, July 26, 2006  
Anonymous Anonymous said...

"SO it's okay for Core 2 at $300 to be faster AT ANYTHING than a $800 and it's ownage, but if AMD does it, it doesn't matter because it's not every test?

And remember the X2 3800+ is now $150."

Is that what I said? All I've been saying is Sharikou's continued comments that the 3800 frags a 965XE is crap and the link he keeps providing doesn't support his statement.

As for your comment above, I'm not exactly sure what you are trying to say. My point is if Sharikou's definition of frag is the 3800 results vs 965XE than a Core 2 is no contest with FX62. (Personally, I don't think either case is true)

Do you honestly believe the link Shraikou keeps providing demonstrates that the 3800 "frags" a 965XE?

1:33 PM, July 26, 2006  
Blogger Sharikou, Ph. D said...

As for the proof that X2 3800+ frags Pentium XE 965, why don't you go over the benches and tally the scores?

1:51 PM, July 26, 2006  
Anonymous Anonymous said...

"And 2008 is when the pain really gets there for AMD. Nehalem. For AMD, that's spelled "D O O M"."

Hahaha! Thank you Dr. Who, I'm sure you know everything that is going to happen in the future. Prophecy for me oh seer!

Seriously if you think Nehalem is going to be AMD's doom, you're smoking some serious dope. Nobody knows exactly how Nehalem or K10 are going to perform. Any belief in either "0wnin" the other is pure fanboy and marketing hype at this point in time.

"People think AMD is feeling the pain now, what will they do when every single Intel technology is better than every single AMD tech? Right now, AMD has HT which is good stuff, and IMC but CSI is better and Nehalem will have IMC and CSI."

CSI will NEVER measure up to HT at it's 3rd and especially 4th generation, go read up the documents from the Hypertransport Consortium.

http://www.hypertransport.org/

HT 3.0's 41.6 GB/s is moving a lot of data and I highly doubt CSI can touch it. CSI = asynchronous, alternating, multi-channel bus, HT = parallel, synchronous, point to point bus. Granted CSI is a step up from thier typical FSB designs but HT is already shaping up to surpass it by the time it already is out.

http://www.theinquirer.net/default.aspx?article=28298

If Intel is so great with an intergrated memory controller, then why haven't we seen them put it to silicon yet?

We'll see just how fun the growing pains are for Intel trying to intergrate a memory controller into the CPU are for them. Even AMD's IMC wasn't the greatest on the initial Clawhammer/Hammer cores.

Considering Intel's current track record, they better pray they get it right the first time with CSI and thier IMC, the K8L is already shaping up to be far more than a match for Conroe and the K10 in 2008 I expect AMD will engineer it to be superior to Intel's offerings at that time. Contrary to your popular belief, AMD is not resting on thier laurels. To think otherwise is a foolish man's thought.

1:57 PM, July 26, 2006  
Anonymous Anonymous said...

AMD a company of greedy cry babies.

First advertise benchmark leadership and sucker the enthusiast to spend 1000 on their CPU.

Next month pooh pah benchmarks and drop that same CPU by 50%.

Gouge the consumer when you can. Firesale when your product looks like garbage.

What a company... what is more funny is the suckers who somehow think the company is more holy then INTEL. They are both run by greedy CEOs and driven by the same primeval urge all companies are driven by; profits, growth, survival.

Too bad AMD doesn't look good for any of those.

The Doctor

4:49 PM, July 26, 2006  
Anonymous Anonymous said...

"As for the proof that X2 3800+ frags Pentium XE 965, why don't you go over the benches and tally the scores?"

You know, I took your excellent advice and decide to tally the scores - and I must admit you were absolutely correct!

The 965 outperformed the 3800+ in 24 benchmarks and the 3800+ outperformed the 965 in 11 benchmarks. I could be off by 1 or two but the I think the #'s are clear.

So if "frags" means losing more benchmarks than winning, you are indeed absolutely correct - the 3800+ "frags" the 965.......my bad!

4:51 PM, July 26, 2006  
Anonymous Edward said...

"If Intel is so great with an intergrated memory controller, then why haven't we seen them put it to silicon yet?"

Intel HAD put IMC to silicon, which is Itanium 2. I'd rather believe that it didn't put IMC into x86 chips now for Itanium's sake. Intel is a big company; on x86 development itself Intel's force is 4x bigger than AMD's. You typically see Intel making progress in all directions - it was developing dual core when hyping hyperthreading; it was developing Conroe/Woodcrest when making p4. All those things were accomplished by different teams in parallel.

I won't be surprised that some part of Intel come up with a nice HT competitor for x86, earlier than we all expect. IMO, the advantage of HT against CSI will not be as much performance as its open and standardized approach.

The desire to control standards to eliminate competition is part of Intel's genes, from PCI to USB to the next generation wireless technology. However, HT and AMD's open architecture approach may (hopefully) finally turn the odds around. that. If HT is accepted quickly enough by a chain of companies around AMD's x86 solutions, Intel will have a very hard time to compete. Either Intel changes and loses its monopolistic power, or Intel doesn't change and (hopefully) loses the competition.

4:53 PM, July 26, 2006  
Anonymous Edward said...

"First advertise benchmark leadership and sucker the enthusiast to spend 1000 on their CPU.

Next month pooh pah benchmarks and drop that same CPU by 50%.
"

Oh yes, when AMD's chips were sold $1000 they were sold like hot cakes. Those X2 3800+ at $300 were so popular that you saw recommendation on every enthusiast forum. AMD made really good products.

Then Intel showed off its Conroe which isn't available for 6+ months. Even at the end of this year, only 20% lucky souls can buy one. So, to attract the general public away from precious Conroes, Intel decided to lower its p4 prices like no tomorrow. Suddenly AMD's chips are relatively expensive! That NOT because they were expensive in the first place, but because its competitor cut prices like a suicide bumber.

So AMD slashes its own prices, too, right before Conroe's launch. Tell me what's greedy about that? Would you rather see AMD not slashing prices and observe falling marketshare? Or do you think AMD should sell all its high-end chips the same price like crappy & (literally) hot P4EE? You must be an Intel shareholder, aren't you?

I personally bought an X2 3800+ for $300 in January. To me, that's nothiing but a great deal. My server runs cool, quiet, fast, reliable, and smooth. I feel lucky that I didn't buy HyperTransport, nor Pentium-D, nor waiting 6+ months for a Conroe, which reportedly runs hotter than benchmarked and offers me risk to mess up my RAID setting.

6:18 PM, July 26, 2006  
Anonymous Anonymous said...

Im getting sick of Doctor's Hypocresy and unhelpful comments.
like "if its maded by intel and if intel people says so, then its good"
but he of course, goes to the totally inversed when AMD or supporters of AMD says something..

PATETIC...


like claiming AMD is for greedy cry babies..
last time I see, intel was a MONOPOLY using its power to mandate trought the entire computer industry, along with Microsoft.

6:35 PM, July 26, 2006  
Anonymous Wirmish said...

INTEL 386 DX-33 < AMD 286 DX-40
INTEL 486 DX4-100 < AMD 486 DX4-120
INTEL Pentium 120MHz = K5 90MHz

0.25µ : INTEL P3 550MHz MAX vs AMD K7 700 MHz MAX

INTEL P3 SDR Bus vs AMD K7 EV6 DDR Bus

INTEL P4... Mouhhhaaaaaa....

First INTEL Win = C2D vs K8.
Now wait for the AMD K8L in 2007.

6:47 PM, July 26, 2006  
Anonymous Anonymous said...

Intel does have a replacement for the FSB. It's called CSI (Common System Interface) and it's due in 2008. It'll be, presumably, used in high end Xeon processors to start with.

7:12 PM, July 26, 2006  
Anonymous Anonymous said...

With the new pricing, it is so sweet deal to get X2 3800.

Sadly, my company just bought one PD805 (the slowest of Netburst dual-core) and one single-core AM2. You know what, within 3 to 4 hours, the PD805 BSOD. Sometimes even worse, the screen shows gibberish pattern in multitude of colors. We send in the PC for fix the 4th time, it is still the same.

Both using same parts except the board and CPU. We even exchange the parts to check for defects.

10:43 PM, July 26, 2006  
Anonymous Anonymous said...

OK!!! Number 1: Intel will still be the dead company in the end
#2: AMD will suddenly beat Conroe next year
#3: Yes, Conroe is more quick, but has horrible stability(imagine alienware conroes!!!)
Yes, the worlds all happy again:)

11:12 PM, July 26, 2006  
Blogger duploxxx said...

edward do you really believe what you wrote?

Intel is a big company; on x86 development itself Intel's force is 4x bigger than AMD's. You typically see Intel making progress in all directions - it was developing dual core when hyping hyperthreading; it was developing Conroe/Woodcrest when making p4. All those things were accomplished by different teams in parallel.

it was one team that developed the core design for mobile and changed it to conroe and woodcrest... one thing we can already say... merom has better performance then yonah has some 64bit implementation and produces much more heat then the yonah. conroe is a nice design in single thread and 32bit and woodcrest clock/clock equal performance to opteron.

all 3 are based on core design you can see that on the powerconsumption. and the lack of decent multithreading on all the core products...
http://techreport.com/etc/2006q2/woodcrest/index.x?pg=6

so forget you're so called multi design team, it was just an extension they made of an old design, the whole r&d of intel is not capable of bringing a new design to the world... and develop when making p4 hello how many years are intel guys stuck with the p4 design?

12:20 AM, July 27, 2006  
Anonymous Anonymous said...

"The desire to control standards to eliminate competition is part of Intel's genes, from PCI to USB"

Those are open standards! And would you say that we were actually WORSE off with PCI as compared to EISA (which was hardly any better than ISA), VL-Bus (which was a horrible kludge) or MCA (which was a real example of what you're talking about)?

1:58 AM, July 27, 2006  
Anonymous Anonymous said...

To the "doctor",

You are such an thick headed idiot!!!
All you do is slag AMD off and saying how bad they are, but deep inside you know that ur ass gets creamed, cause Intel is in trouble with their crappy P4's and unable to deliver the new corn2row.
Even if they have the faster cpu, who cares, do you honestly think people notice the difference between a X25000 or a conroe...I don't think so.
Maybe if we were running superpi all day, yeah we would, but unfortunately for you we don't!!
So you just buy your conroe and we just keep buying our amd's and we are all happy.
And stop slagging Sharikou, he might be pro AMD, but atleast he backs up his stories...all you can do is say he is an idiot bla bla bla

Get a life!!

The Nurse

2:13 AM, July 27, 2006  
Anonymous Anonymous said...

btw, shari..
found this link..
its dedicated to that retard who claimed the Conroe's 64 bit instructions are better than AMD's


http://www.theinquirer.net/default.aspx?article=33309

12:26 PM, July 27, 2006  
Anonymous Anonymous said...

http://www.theinquirer.net/default.aspx?article=33307

and that conroe seems to be a paper launch despite having 10 suposed million worth of conroes out.

12:42 PM, July 27, 2006  
Anonymous Edward said...

"Those are open standards! And would you say that we were actually WORSE off with PCI as compared to EISA (which was hardly any better than ISA), VL-Bus (which was a horrible kludge) or MCA (which was a real example of what you're talking about)?"

No, I didn't mean the ones you mentioned are beter, but this is beside my point, which says that the main reason Intel backed PCI-X/e was, well, these were designed by Intel. Other industry standards, such as HyperTransport, IEEE-1394, etc., are probably superior to Intel's versions, but Intel treated them as non-existent.

It is this type of attitudes towards competing standards that I was critisizing about. They are bad for end consumers and for the industry.

3:31 PM, July 27, 2006  
Anonymous Anonymous said...

"btw, shari..
found this link..
its dedicated to that retard who claimed the Conroe's 64 bit instructions are better than AMD's

http://www.theinquirer.net/default.aspx?article=33309"

To the idiot who posted the link above, did you actually look at the link in the Inquirer article, it showed Conroe beating FX62 in all but one of the 64BIT BENCHMARKS! I sent a note to the idiot at the inquirer who wrote the article asking him about his conclusion and his response was:

'The title is "AMD does 64-bit better than Conroe". That is open to
interpretation and is rather an apetizer for you to read what Xbitlabs has written. I am someone who review reviews rather than writing them myself.'

His email is desire.athow@gmail.com, and I can forward the note for folks who don't believe me.

9:34 PM, July 27, 2006  
Anonymous Anonymous said...

"Other industry standards, such as HyperTransport, IEEE-1394, etc., are probably superior to Intel's versions, but Intel treated them as non-existent.

1394 was really done in by Apple's stupidity more than anything else. Intel were actually going to include a 1394 controller in their chipsets at about the time the Pentium III came out, but like most other hardware makers, they wouldn't pay the ridiculous licensing fees Apple were asking for. By the time Apple reduced them to a reasonable level we were on to USB2, and 1394's window of opportunity had been slammed shut.

10:54 PM, July 27, 2006  
Anonymous Wirmish said...

Intel PCI-eXpress vs Good Old AGP... 8þ

http://www.hkepc.com/hwdb/gf7600gsagp-2.htm

Thank you Intel...
Good innovation...
Super hyper cool technology...
What's your next big thing ?

Have to scrap my 'old' APG motherboard now...
Thank you very much buddy...

10:17 AM, July 28, 2006  

Post a Comment

Links to this post:

Create a Link

<< Home