.
Quiksilver
Diamond Member
- Jul 3, 2005
- 4,725
- 0
- 71
Well you can get off AMD bandwagon because of price/performance, but to abandon AMD because some foreign company owns 8.1% of AMD stock is not really a good reason. In a global economy today, I wouldn't be surprised if some Saudi royal family owns a few percent of Intel too. Many of them are the largest shareholder of some of the largest companies in the US.
I am disappointed in the Phenom launch like everyone else. But I disagree with all these doom and gloom prediction of AMD. We enthusiast may think we are THE market, but that's really not true. Both AMD and Intel relies much more on their OEM partner for revenue. AMD is in a much better position with OEM like HP, Dell, Toshiba and many others then back in the dog days. So they will sell their cpus. The only problem is they got too big headed with their A64 success and bought ATI, now they really have to trim their expenses to get back to profitability or just get even, that's the key to the survival of the company.
As I see it, they will do everything they can to survive for now, get cash from foreigners to issue more bond and sell fab or even spin off ATI. And I hope with this lesson, AMD will learn to focus on their main product, CPU, and focus on engineering team rather than PR people. Eventually they will come up with another solid product like A64.
I think you are massively underestimating the complexity of the devices we are talking about here.
Designing a new microarchitecture is an exercise of enormous complexity and difficulty. There are lots and lots of factors at play, many of which are difficult or impossible to control, and because of the long-term nature of the project many decisions must be made based on the designers' intuition of what the world will look like in 5 years.
There are factors that are warning signs that overall scalability or performance may not be as desired (the former can be quantified with simulations, to an extent), but these are subtle and even very smart people can still misread the situation. A modern CPU has lots and lots of components, and these can interact in ways that are difficult to adequately predict. Scalability is also heavily impacted by the process technology, which was not in place when development began.
Consider that on the Intel side of the fence, both the Netburst and Itanium teams severely overestimated their scalability at the outset. There's a famous quote about how Netburst had headroom to hit 10 GHz; it never quite reached 4 GHz. At the outset of the Merced project, there was very serious talk that Itanium would completely replace x86 by quickly ramping up clock speeds, yet this did not happen. Conversely, consider that the P6 project knew that their first product (the Pentium Pro) would not make a lot of money (it was a horrible nightmare to manufacture, so volumes were low), and yet their underlying design was spectacularly successful.
Barcelona is not off to an especially auspicious start. But then, Willamette did not exactly turn heads either. A few years later, the Northwood "C" products certainly faired much better. The doomsaying of how Barcelona will destroy AMD is very premature.
colonel
Golden Member
- Apr 22, 2001
- 1,784
- 21
- 81
I buy AMD only because I remember stuck, i Pentium II with poor performance until the Athlon slot A showed up and blew the whole monopoly of Wintel Windows+Intel for the half of money of the Celeron then.Everything in America without delay is from China o somewhere so it makes your statement useless.
JonB
Platinum Member
I'm not a gamer. I don't care if my video frame rates are 120fps instead of 110fps. I do know that my AMD64 3000+ and my Opteron 165 are still strong performers for what I do 99% of the time.
Will I replace one of these with a Barcelona and AM2 mobo? Very likely. The performance difference between XP and Vista is much more of a concern to me.
Will I replace one of these with a Barcelona and AM2 mobo? Very likely. The performance difference between XP and Vista is much more of a concern to me.
mindless1
Diamond Member
- Aug 11, 2001
- 8,733
- 1,746
- 126
A going-away party is as good an excuse as any to throw a party. Woohoo, BYO :beer:.
I know what you mean though ebeattie, I feel exactly the same way about Cuba. Not the people, I mean it's just so suspicious to have it sitting there in the middle of a gulf, an ocean and a sea, like the whole world revolves around it, and it's been that way for far too long. I will no longer look at Cuba on a map and it will be sorry for offending me! Don't even get me started about Cheerios. :evil:
All kidding aside, if you keep your identity a secret, I think you have a good shot at getting a vendor to sell you an Intel CPU and we won't tell.
bryanW1995
Lifer
- May 22, 2007
- 11,144
- 32
- 91
myocardia
Diamond Member
- Jun 21, 2003
- 9,291
- 30
- 91
I'm pretty sure he meant PIII. But, the first Athlon debuted in June of 1999, and the Pentium 2 debuted in May of 1997. Oh, and the first Athlons were about as much faster than the PIII's as the C2D's are the Athlon X2's.
Wonder what the OP is going to do when he has to drive to work in his assembled in mexico car, using oil from the mid-east, and watching TV on his chinese/tiawanese set.
All sarcasm, aside, the only thing positive he has done, is mabye learned a lesson about blindly supporting a product aka fanboism, but i doubt it.
Supporting amd/intel/ati/nvidia or whoever is fine, but the simple fact is you should support whoever makes the best product for the ammount of money you will spend.
All sarcasm, aside, the only thing positive he has done, is mabye learned a lesson about blindly supporting a product aka fanboism, but i doubt it.
Supporting amd/intel/ati/nvidia or whoever is fine, but the simple fact is you should support whoever makes the best product for the ammount of money you will spend.
I'm not exactly happy with what AMD has put out with the current Phenoms, but I'm probably going to buy one next year when (hopefully) faster models are out in Q1 and 45nm models are out in Q2-Q3. I already have an AM2 system with an X2 3600+ that has provided me with excellent value. Sure, Intel has higher performance, but with AMD I can just buy an X4 right now or even a 45nm model next year and drop it into my AM2 system. Hopefully next year we'll be able to buy a $200-250 quad core that will overclock to 3GHz or so. If AMD has a product like that I'll definately buy it.
Difficult to calculate for microprocessors alone, comparing each company together with all the products they make as well as using the basis that if Intel + AMD = 100% revenue share then AMD represents ~14% for Q3 2007.
I have always been a big AMD fan the last few years since P2 came out but since the AM2 boards I have been thinking about switching. I think made a plunge and bought a C2D a few months ago and was wondering why I didn't switch before, the board layouts are better and the processor just seems more stable. I feel like I can't overclock quite as much on the Intel but it just runs cooler. I have felt the same about Nvidia vs ATI, been a fan of ATI for a while but lately it's been like they have been lacking with their drivers on the new games so I almost want to switch to Nvidia. I work around a lot of Intel junkies and they always look at me wrong when I tell people that they should buy an amd processor, they really are a great value for the money. I wouldn't exactly use them for hardcore gaming but for just the casual user they aren't so bad at all.
Hulk
Diamond Member
- Oct 9, 1999
- 5,118
- 3,664
- 136
Originally posted by: Viditor
I'd like to clear up a few common misconceptions here if I may...
I have seen these time and again, and if you don't really study the semiconductor industry as I have, it's easy to arrive at them by using "common sense". Unfortunately, they're wrong...
1. "AMD got lazy" - A very common misconception... The fact is that it takes ~5 years and a huge amount of money to create a newly designed chip. It IS possible to cut this time down a bit (say 3.5 years) if you throw a ton of extra money at it (double or triple the design teams while allocating plenty of extra Fab lines to R&D for example).
But for AMD to have a killer replacement of the X2 in place by 2007, they would have had to start it in 2002...
How many of you remember what happened to the economy in general and the semiconductor industry specifically in 2002? That's right...disaster to the bottom line for both Intel and AMD! Intel's share price dropped from ~$35 in Jan 02 to ~$13 in Oct 02, and AMD went from ~$20 to ~$3.50...Both companies had their cash hurt badly because nobody could afford to buy computers anymore.
Of course for Intel, this wasn't such a big deal...they still had many Billions in reserve, stored up over a number of years. But for AMD, launching a new large-scale R&D effort was like pushing a boulder uphill in the dead of winter during a blizzard, while dragging a sled and all the dogs...it was a bit tough.
Now add to that...When Prescott was nearing completion (probably when it first taped out and they could test it, say early 2003), Intel realized that they were about to hit a HUGE wall very quickly. The leakage in Netburst had become much greater than anyone at Intel could have predicted...
So, Intel made the smartest strategic move that they have in their history. They bit the bullet, cancelled a whole slough of projects (like Whitefield, which was to be the first CSI based CPU to be released in 2006/7), and poured ALL of their resources into their existing Core and Core2 projects...the upshot was, they got C2D out in 3.5 years instead of the usual 5 years and there was absolutely NOTHING AMD could do about it...they just couldn't afford those kind of resources.
BTW...I find it quite ironic that many folks feel that AMD is both lazy, and that they made a mistake investing all that money into ATI.
Buying ATI is in essence AMD being highly motivated to not be behind for the next great innovation. In fact they hope to be well out in front (though it sure is painful and expensive to get there...).
2. "AMD has fallen way behind in Marketshare because of C2D" - In point of fact, AMD's marketshare is much higher than it was before C2D.
AMD's current marketshare is 23.5%, prior to C2D it was 21.6% or less...
Great post but I don't understand what you are implying with the last point, #2? It seems that you are implying that the release of C2D was a GOOD thing for AMD since their market share actually increased since the release? I very much doubt that.
I think there is a lag in the market from when a chip is released and when it actually has an effect on market penetration and numbers and the statistics you quote are reflecting the lag of C2D market penetration.
I also believe the poster you are responding to is correct in that C2D was a huge hit to AMD. Citing a statistic that seems to negate that fact doesn't change the truth of the matter and that C2D IS the reason AMD is struggling. If Intel had continued on with the Netburst and there was no C2D AMD would be flying high right now.
WTF is about all these lazy comments? They were forced / made a dismal ballsy gamble, to go monolithic quad core. Thats it. Done. Plain and simple. Damn it! If they went MCM like the smart money did, we wouldn't be having this discussion as AMD Quad cores would've been out in droves 6-12 months ago. They had to make a huge core work on a 65nm process. Intel did not. As I explain in another thread, the majority of phenom's problems (lateness, apparent lack of scaling in early spins, etc) stem from that issue alone. HUGE monolithic core, and no money for a die shrink. In fact, can anyone give me a larger core right now then Phenom in this market segment? How about designed and produced by a company with AMDs Market cap? Any takers?
Separate the issues here. Bad gamble = much harder work to achieve the status quo. Phenom isn't even working right still, and thats a testament to an in ability to focus the still mostly separate groups that compose AMD/ATI right now. Also, notice the leadership is slowly being lead to the guillotine? AMD didn't get lazy, they had to pull off something Intel (10x the size) wouldn't touch. With 2 fabs. And a small fraction of the money...
It was all or nothing, they couldn't go sh*t we screwed up here's a billion to go tapeout and mfg MCM x2s, if they had the resources they very well might have... But they didn't.
I understand the frustration, I too am frustrated by an inability to execute... But attribute it to the right things, not foreign ownership and lazyness. But lack of resources, and capacity ramping ( without which we'd be saying they were stupid for not taking advantage of the window where they had the superior product... shareholders sue over that ).
The seeming increase in market share, is due to mentioned capacity ramping... AMD sells almost every single cpu it makes, it was capacity holding back market share, not inability to sell product. The C2D reference with market share is not indicating ( i think) that c2d helped AMD, simply that it didn't effect quantity sold, only ASP.
Separate the issues here. Bad gamble = much harder work to achieve the status quo. Phenom isn't even working right still, and thats a testament to an in ability to focus the still mostly separate groups that compose AMD/ATI right now. Also, notice the leadership is slowly being lead to the guillotine? AMD didn't get lazy, they had to pull off something Intel (10x the size) wouldn't touch. With 2 fabs. And a small fraction of the money...
It was all or nothing, they couldn't go sh*t we screwed up here's a billion to go tapeout and mfg MCM x2s, if they had the resources they very well might have... But they didn't.
I understand the frustration, I too am frustrated by an inability to execute... But attribute it to the right things, not foreign ownership and lazyness. But lack of resources, and capacity ramping ( without which we'd be saying they were stupid for not taking advantage of the window where they had the superior product... shareholders sue over that ).
The seeming increase in market share, is due to mentioned capacity ramping... AMD sells almost every single cpu it makes, it was capacity holding back market share, not inability to sell product. The C2D reference with market share is not indicating ( i think) that c2d helped AMD, simply that it didn't effect quantity sold, only ASP.
Listen pal, the only person you're delusioning is yourself. You wouldn't have the performance offerings you have now were it not for the athlon64. Btw, did you know the core 2 architecture was a continuing developement of the their P3 which became centrino then became core then core 2? Did you know that the continued developement has all been done by a firm aquisitioned by intel years ago in a little country called israel. You're wanting to be loyal to them now because they're more american? Open your eyes and read more. If you think intel is so much better, go ahead and jump ship and leave your amd system to charity. Good luck with backwards compatibility, enjoy spending that money twice and 3 times when you want the latest from intel, cause in a couple months the new cpus and chipsets from intel come out and none of the current stuff will work with it.
I'll reply to this one too. Maybe you didn't hear about intels big money loss in the ram industry they invested in and are folding shop on still. Maybe you didn't hear about the nearly 1 billion dollars they paid out in japan and europe for using strong arm tactics to keep their p4's floating on top, where most things like it float in the toilet bowl. AMD is a much smaller company and they're shooting for a bigger picture than immediate cpu performance that will blow the socks off the competition. That may not be the smartest approach short term, but lets see how long intel can keep their product looking rosey in the long term. When IT people start looking at long term (5 years or more) backwards compatibility of future products, alot of them will look at amd first. Case in point all socket f opteron servers can drop a barcelona quad core in with a bios flash and that sever will still be a viable sever when it hits age 5 or more. You can't say that about any of the xeons that are just 1.5 years old with netburst xeons in them. You want better performance from them you have to completely replace them with new stuff, and if you're running a datacenter with 300 rack mount servers, the cost is clearly in amd's favor. You need to take a chill pill and read a little more closely before you make such assenine posts.
I really have no idea what you're expectations are or perceptions are, but hey line up like a sheep and join the it's popular to change sides constantly band wagon. I'm sure we'll see a near identical post from you when amd finally gets back on top again and it will happen. Oh? You say won't happen eh? People were saying amd would never be on top when the athlon xp was the best they offered and intel was scrambling to stay ahead of it. Oh wow then the athlon64 trounced intel for 3 years running, but the haters said that would never happen, just like they're saying it will never happen again. Key difference they had to implement the word again in that statement.
@zsdersw: again... Sigh... There would be no C2D this soon, without K8 X2.... Sigh... and yes Everyone and their dog with half a brain knows C2D is a revision of the multiple cores that was the Core project, which was a rethink on teh P3, which was closely related to the P2, which was itself a re-adaptation of Ppro (aka P6)... Which had some slight resemblance to the P5 (pentium) which differed massively from the 4x86... Much like Athlon was a massive departure from K6/x... And yes.. both C2d and Athlon architectures really just follow in the design footprints of Alpha... Are we happy now?
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter