AMD vs Intel vs Apple (NO FLAMING!!!!)

[Lord Kalthorn]

Pentium M is going into a lot of the mainstream Desktop ones too It is great; although I'm not a power person myself - I don't care whether my chips take 90 or 130 or 200. As long as I can find a PSU for it, and a cooling system for its its fine with me.

Haha; I refuse to accept what is not yet certain.

Haha; saying let me explain and then copying from a Paper on the subject without Quotes or saying where is comes from... so silly Especially when you don't get rid of evidence it wasn't written by you or for this synax with many tensual problems to the situation. It would be interesting to hear where this comes from, and a link to the rest.

However this is a very interesting article. To however presume both AMD will carry this all out; and that Intel is not doing it as well without making a big point about it because its an anal detail that nobody needs to know. Few people need to know how the processor does what it does; a good number do but most don't and they do not need more branding rammed down their throat. Pentium 4 HT EE is already enough to swallow. AMD Athlon Toledo HS SMT is just a useless mouthfull, and they probably will include it.

Yeah; so give us a link Or is this from a White Paper that is not publically viewable?

 

[DarkBlade1]

Umm, LK, I didn't copy and paste jack shit. Don't accuse me of that please. That was a paper I wrote for advanced physics on Ultrascalar technologies that I presented at the Junior Academy of Science for the Arts and Humanities over this past Summer. Don't try to make me look cheap and half rate, THX.

Intel will be going Ultrascalar as well, but they will be using CSL (close to the speed of light) optical die interconnects to do this while AMD will stick with their SOI (Silicon on Interconnect) Intel's OOI will be out 2007, and AMD may make the move also, but no one knows.

 

[DarkBlade1]

Here, I can take more from my paper if you want me to.

The Ultrascalar processor core performs the same functions
as a typical superscalar processor core. It renames registers, analyses register and memory data dependencies, executes instructions out of order, forwards results, efficiently reverts from mispredictions, and commits and retires instructions. The Ultrascalar processor core is much more regular and has lower asymptotic critical-path length than
todays superscalars, however. In fact, all the scaling circuits within the processor core are instances of a single algorithm, parallel prefix, implemented in VLSI. Because of the core's simplicity, it is easily apparent how the number of gates within a critical path grows with the issue width and window size. The core does not include the memory and branch prediction subsystems. Instead the core presents the same interface to the instruction fetch unit and the data cache as today's superscalar processor cores. Predicted instruction sequences enter the core, and data load and store requests are initiated by the core. The Ultrascalar core will benefit from the any advances in effective instruction fetch rate and in data memory bandwidth that can be applied to traditional superscalar processors. In particular, since the Ultrascalar processor core performs the same functions as the core of today's superscalars, it achieves the same CPI performance as existing superscalars when attached to a traditional 4-instruction-wide fetch unit using traditional branch prediction techniques and a traditional cache organization. As effective fetch rates and data and widths increase, the Ultrascalar core can scale gracefully, raising the CPI without exploding the cycle time.

I did this paper in conjunction with Yale University and many professors sponsoring my project. I placed 3rd in the region with it. I was very pleased, but should have gotten first place. Ultrascalar abilities are heavily impeded by CAS Latency on current RAM developments, and so, a CAS Latency of 2 nanoseconds or lower is required to complete the Logarithmic functions correctly under the scale the clock cycles are operating in.

 

[Lord Kalthorn]

Umm, LK, I didn't copy and paste jack shit. Don't accuse me of that please. That was a paper I wrote for advanced physics on Ultrascalar technologies that I presented at the Junior Academy of Science for the Arts and Humanities over this past Summer. Don't try to make me look cheap and half rate, THX.

Intel will be going Ultrascalar as well, but they will be using CSL (close to the speed of light) optical die interconnects to do this while AMD will stick with their SOI (Silicon on Interconnect) Intel's OOI will be out 2007, and AMD may make the move also, but no one knows.

Then you did copy and paste it? Just from yourself instead of somebody else. As you wrote it you should have the knowledge to write it in context again for our consumption. Why have you got a bit in a paper from Summer about Cores whose names were not released in the Summer?... You don't normally write like that Its strange.

Haha; so what you're saying is AMD are using Silicon; and Intel are using Close to the Speed of Light Optical Die? One sounds faster to me; and its not the fore. You're also saying that AMD will make the move after Intel now?

Have you changed teams AMD?

 

[DarkBlade1]

I added core names after I copied and pasted from the paper. I didn't know those core names at the time, duh! The technical parts of the paper are copied from the paper on my PC, the addins are the comments about the companies. The paper included nothing about companies, it was merely about scaling on CPUs.

No, haven't changed teams, Intel is breaking past AMD in raw power for now, and the 65nm process is REALLY COOL. AMD hasn't released statistics on their 2006 plans really, so who knows... Intel are smart, I am interested to see what they can pull off. Right now, power consumption is their biggest issue. If that wasn't as high, I would give them more credit.

 

[Lord Kalthorn]

Here, I can take more from my paper if you want me to.

The Ultrascalar processor core performs the same functions
as a typical superscalar processor core. It renames registers, analyses register and memory data dependencies, executes instructions out of order, forwards results, efficiently reverts from mispredictions, and commits and retires instructions. The Ultrascalar processor core is much more regular and has lower asymptotic critical-path length than
todays superscalars, however. In fact, all the scaling circuits within the processor core are instances of a single algorithm, parallel prefix, implemented in VLSI. Because of the core's simplicity, it is easily apparent how the number of gates within a critical path grows with the issue width and window size. The core does not include the memory and branch prediction subsystems. Instead the core presents the same interface to the instruction fetch unit and the data cache as today's superscalar processor cores. Predicted instruction sequences enter the core, and data load and store requests are initiated by the core. The Ultrascalar core will benefit from the any advances in effective instruction fetch rate and in data memory bandwidth that can be applied to traditional superscalar processors. In particular, since the Ultrascalar processor core performs the same functions as the core of today's superscalars, it achieves the same CPI performance as existing superscalars when attached to a traditional 4-instruction-wide fetch unit using traditional branch prediction techniques and a traditional cache organization. As effective fetch rates and data and widths increase, the Ultrascalar core can scale gracefully, raising the CPI without exploding the cycle time.

I did this paper in conjunction with Yale University and many professors sponsoring my project. I placed 3rd in the region with it. I was very pleased, but should have gotten first place. Ultrascalar abilities are heavily impeded by CAS Latency on current RAM developments, and so, a CAS Latency of 2 nanoseconds or lower is required to complete the Logarithmic functions correctly under the scale the clock cycles are operating in.

Ah see now that is information! It does sound interesting... luckily however I'm a software person As long as it runs Visual Studio and works I don't need to know why Thank god there are people to do that like yourself for us.

I added core names after I copied and pasted from the paper. I didn't know those core names at the time, duh! The technical parts of the paper are copied from the paper on my PC, the addins are the comments about the companies. The paper included nothing about companies, it was merely about scaling on CPUs.

No, haven't changed teams, Intel is breaking past AMD in raw power for now, and the 65nm process is REALLY COOL. AMD hasn't released statistics on their 2006 plans really, so who knows... Intel are smart, I am interested to see what they can pull off. Right now, power consumption is their biggest issue. If that wasn't as high, I would give them more credit.

Raw power is all that matters is it not? The 65nm process looks cool even from my limited perspective on the subject. Although HPs Molecular Processor Strings look far more interesting If not a little more in the future.

That would be useful! Why aren't they... silly buggers. Maybe they don't have any plans. They just registered a bunch of random names for the publicity and now have us talking about them without even saying they exist! 130 whatevers, probably Watts, doesn't sound too bad to me?

 

[DarkBlade1]

Lord Kalthorn said:
Ah see now that is information! It does sound interesting... luckily however I'm a software person As long as it runs Visual Studio and works I don't need to know why Thank god there are people to do that like yourself for us.
Raw power is all that matters is it not? The 65nm process looks cool even from my limited perspective on the subject. Although HPs Molecular Processor Strings look far more interesting If not a little more in the future.
That would be useful! Why aren't they... silly buggers. Maybe they don't have any plans. They just registered a bunch of random names for the publicity and now have us talking about them without even saying they exist! 130 whatevers, probably Watts, doesn't sound too bad to me?



Well, look at it this way. The P4 EE 3.46GHz use around 116 Watts, and look how much heat it produces!! Now, it is a fact that the more voltages and wattage put into the internal die of a processor, the more heat it will lose to entropy. With that in mind, image 33% more heat in the next P4 series.... That will be enormous, and it will require the use of the less effiecient Si8 (silicon with 8 bonds in a chain) Currently, Si4 is used in a lattice to make the wafers on die. It is very efficient (AMD uses Si3, a strange more energy effiecient variation chemically synthesized in the lab), while Intel uses the Si4 lattice to their work. Both are efficient, but Si3 uses less energy to entropy conversion, but cannot bear as much heat, and thus, AMD is unable to push their clock rates as high without chemical additives. The, 3800+. 4000+ and FX-53/55 use an additive in the Silicon wafers (I believe it is Si02 (also known as a form of glass or quartz) to assist in attaining higher clocks. Too much of the quartz like material can increase conductivity, thus ruining the die and wafers. So, AMD has to be very careful with what they do. Intel is going the wrong path here, shooting for higher and higher speeds , when they should focus on more efficiency.

 

[rakedog]

Stop using colors AMD...............

Thought you're pretty much right about your information.

 

[DarkBlade1]

What is wrong with colors? They make the forum look better! I mean, lets add some life into this place, we are past 100K in posts, lets act like it!

 

[rakedog]

No. Colors make a forum look stupid. No big forum uses colors, it's just annoying and tough on the eyes. Don't use colors, period.