Is this going to be a penny stock? I can't see INT being the only player out there. It going to be at a 5 year low pretty soon.
AMD is it playable at this level?
- Thread starter lowhangingfruit
- Start date
My how times have changed. Just about 3 years ago, AMD had Intel at the tipping point. All they had to do is to give them one more push and AMD would have started the avalanche to supremacy.
AMD has no cash. Where are they going to get the funding from to open the plant? They are leveraged to the max.
sigh.
AMD is not going down. That would give Intel a monopoly.
AMD and Intel are separated by an overblown (but marginally significant) 20 nanometers, in the grand scale of things.
That 20 nanometers lead will disappear within the next several months.
AMD is a buy and hold all the way. They'll catch up to Intel and pass them yet again. Especially now that IBM is throwing tech support behind them in NY.
It's all like so many phases of the moon.
AMD is not going down. That would give Intel a monopoly.
AMD and Intel are separated by an overblown (but marginally significant) 20 nanometers, in the grand scale of things.
That 20 nanometers lead will disappear within the next several months.
AMD is a buy and hold all the way. They'll catch up to Intel and pass them yet again. Especially now that IBM is throwing tech support behind them in NY.
It's all like so many phases of the moon.
Intel and AMD are not separated by just a "20 nanometer" difference. Up until 2004, AMD had a superior architecture. Intel had the CISC instruction set and AMD had a RISC architecture.
Intel with the Pentium IV had serious problems with how their frontside bus was a complete disaster in terms of memory latency. AMD had the on chip memory controller. Intel had real problems with power consumption.
AMD had the superior architecture and the superior processor and were even starting to get long-time Intel supporters like DELL to switch over to AMD chips. AMD had a massive technological advantage over Intel. If they had the cash and the ability to aggressively market their chips, they would have been able to send Intel falling. But they didn't.
And all this time, Intel was creating a whole new architecture. They were building a brand new platform. They then released the Core2Duo and the Centrino platform. This is basically a reusable mobile platform with phenomenal intelligent power management. Along with this, the new architecture gives Intel a large advantage in terms of architectural design. The new platform not only matches AMD in terms of a more intelligent instruction set and on chip controller, but it also provides hyper threading and large memory caches.
So, there is much more to a chip design that just "how many nanometers" a design is. AMD doesn't have the ability to fund its R&D anymore. They have no cash. They had their chance and they blew it. AMD *could* survive, but it won't surpass Intel in technological superiority again until it finds a way to eliminate its cash problem. Without this fix, AMD is not going to be a very successful company for many years to come.
Intel reaped benefits by going back to the Pentium III architecture when engineering the Pentium M, and consequently, Core2 Duo processors, which is what gave them their recent lead over AMD.
They went back to that well in an effort to reduce power consumption, and found that they also reduced a processing bottleneck and stemmed leakage, also.
It was part luck. And it coincided with the much larger growth of laptops versus desktop PCs. They ended up with a processor that used less power, yet was more powerful than their heat-soak monster Pentium 4 that was their prior best effort, but which depended on pure clock speed.
They put the cherry on the whipped cream by dramatically ramping up L2 cache levels, especially in their new Penryn line of processors.
AMD's launch of Barcelona was flawed, as their were specific fabrication defects in the processor. But Barcelona is 65nm, will go 45nm, soon, and had great potential to close the gap with Intel.
Right now, Intel does have a 8 to 12 month lead over AMD because of the Memron and now, Penryn, processors, but AMD may be able to cut that down to something more along the lines of 6 months.
However, Penryn is not exactly the 'revolutionary' product Intel is hyping it as, IMO. The gains by Penryn over Memron are not dramatic, in either greater speed or power consumption. It looks like Intel is stretching (its admitted edge) a little bit here, as it seeks to keep the window with AMD as wide as possible.
32nm is not far away, either.
Regardless, multiple core, along with hafnium and other exotic wafer materials are now the way Moore's Law will sought to be upheld.
Intel got burned badly by failing to realize the importance of 64 bit. They were overconfident. It will happen again (maybe now).
AMD is going to surprise quite a few people with PUMA this summer.
http://www.reghardware.co.uk/2008/03/06/amd_announces_griffin/
The race is far from over.
They went back to that well in an effort to reduce power consumption, and found that they also reduced a processing bottleneck and stemmed leakage, also.
It was part luck. And it coincided with the much larger growth of laptops versus desktop PCs. They ended up with a processor that used less power, yet was more powerful than their heat-soak monster Pentium 4 that was their prior best effort, but which depended on pure clock speed.
They put the cherry on the whipped cream by dramatically ramping up L2 cache levels, especially in their new Penryn line of processors.
AMD's launch of Barcelona was flawed, as their were specific fabrication defects in the processor. But Barcelona is 65nm, will go 45nm, soon, and had great potential to close the gap with Intel.
Right now, Intel does have a 8 to 12 month lead over AMD because of the Memron and now, Penryn, processors, but AMD may be able to cut that down to something more along the lines of 6 months.
However, Penryn is not exactly the 'revolutionary' product Intel is hyping it as, IMO. The gains by Penryn over Memron are not dramatic, in either greater speed or power consumption. It looks like Intel is stretching (its admitted edge) a little bit here, as it seeks to keep the window with AMD as wide as possible.
32nm is not far away, either.
Regardless, multiple core, along with hafnium and other exotic wafer materials are now the way Moore's Law will sought to be upheld.
Intel got burned badly by failing to realize the importance of 64 bit. They were overconfident. It will happen again (maybe now).
AMD is going to surprise quite a few people with PUMA this summer.
http://www.reghardware.co.uk/2008/03/06/amd_announces_griffin/
The race is far from over.
So you agree then that the transistor size is not the only large difference between Intel's new architecture and AMD's? 
I also have to completely disagree that it was "luck" on Intel's part. Intel was fully, fully aware of why the AMD architecuter was superior. They went completely back to the drawing board. They put multiple engineering teams in different locations all in parallel on the task of reengineering for this problem. There was no "luck" in leapfrogging AMD. It was a forceful effort with a lot of money spent on R&D.
I also have to disagree that Puma from AMD is going to leapfrog Intel. It isn't really a superior chip. It is old technology for their cores with some new power management components. Where's the marketing point for buyers? "It's not faster, but it uses less power"? That's a far cry from when AMD could not only brag about being superior in power consumption but also performance.
But anyways, my point is being missed. AMD had the superior architecture FOR OVER A DECADE. They couldn't capitalize on that fact. They couldn't market their chips effectively to the masses, they couldn't produce them efficiently (defects have always been a major problem for them) and they couldn't keep up with Intel's R&D efforts.
What's going to be different this time. They're in worse shape right now financially then they've ever been. I'm sorry, but AMD is more than 12 months behind. *IF* they had the cash, they *might* be 12 months behind (although I think it's more like 24-30 months behind technologically). The fact is that AMD is a horribly managed company that has never been able to take advantage of their advantages. They are a good 3-5 years behind in terms of technology, process and management. Even if AMD improved all three of these problems today, they would still need at least 3 years to fully implement them to catch up to Intel.
I also have to completely disagree that it was "luck" on Intel's part. Intel was fully, fully aware of why the AMD architecuter was superior. They went completely back to the drawing board. They put multiple engineering teams in different locations all in parallel on the task of reengineering for this problem. There was no "luck" in leapfrogging AMD. It was a forceful effort with a lot of money spent on R&D.
I also have to disagree that Puma from AMD is going to leapfrog Intel. It isn't really a superior chip. It is old technology for their cores with some new power management components. Where's the marketing point for buyers? "It's not faster, but it uses less power"? That's a far cry from when AMD could not only brag about being superior in power consumption but also performance.
But anyways, my point is being missed. AMD had the superior architecture FOR OVER A DECADE. They couldn't capitalize on that fact. They couldn't market their chips effectively to the masses, they couldn't produce them efficiently (defects have always been a major problem for them) and they couldn't keep up with Intel's R&D efforts.
What's going to be different this time. They're in worse shape right now financially then they've ever been. I'm sorry, but AMD is more than 12 months behind. *IF* they had the cash, they *might* be 12 months behind (although I think it's more like 24-30 months behind technologically). The fact is that AMD is a horribly managed company that has never been able to take advantage of their advantages. They are a good 3-5 years behind in terms of technology, process and management. Even if AMD improved all three of these problems today, they would still need at least 3 years to fully implement them to catch up to Intel.
btw...Penryn wasn't supposed to be this "revolutionary" product. Their marketing people (which Intel does extremely well: see the clockspeed scam) may have pushed it that way, but Intel always releases an improved version of a similar architecture already released in between times waiting for their next real advancement.
Their new superiority will be in Nehalem which is going to blow the doors off AMD and set them back 2 steps even as AMD makes 1 step forward this year.
Also, for clarification, I am no "Intel fan". 3 years ago, I would have happily bought an AMD because it was superior. I've followed both of these companies closely for a number of years, so I've known both of their weaknesses and advantages. Intel hasn't been this far ahead of AMD since the early 1990's. There is a reason why they've been able to dramatically take the lead and it has to do with AMD and their incompetence as a company. Their "technological" skills have never been questioned, but the company has not been managed well and now they're paying the price. They've been able to keep the gully from flooding by having a superior product for many years, but now the pied piper has caught up with them in terms of their horrible management.