From post above:
There are some decent ideas in the proposal:
- Novelty as the only condition
This is hardly a good idea.
As a matter of fact, with respect to software, novelty is almost guaranteed in that everything you do is just a big set of combinations of the various building blocks. At some level, just about anything could be argued to be novel and has been.
For example, combining internet purchasing, clicking a button and the idea that fewer clicks is better, you end up with the novel (barely) method of 1 click purchasing. Hell, I might as well write a patent generator program that will spew out all combinations of various methods/sequences/actions, cherry pick the ones I think look most valuable and buy my lottery tickets by patenting all of this crap.
By eliminating the non-obvious requirement you have just made an incredibly flawed system (because there isn't enough staff to check up on non-obvious properly) into the worst of all possible worlds.
The only saving grace is that each patent will expire after 4 years.
These are the types of patents that must not be allowed under any patent system:
1) Amazon one-click (gee, who would have thought that fewer clicks would be better)
2) NTP and RIM for e-mail patents (transmitting data over an existing transmission medium???? Are you serious???)
3) System for keeping track of cell call minutes and stopping service when maximum reached
4) Method for distributing nutritional information over mobile devices (you...have...got...to...be...freaking...kidding.. .me. really.)
"I'd never want the FULL version of these databases, even if those, too, were free"
Well, in an enterprise application environment you would want them. In an enterprise application environment, footprint is near the bottom of the list of priorities. The top of the list include:
Reliability
Scalability
Performance
In the 80's and 90's IBM service reps ran around with something they called a "brick" which was a wireless device that they used to communicate with the main office, wouldn't that be prior art for RIM and NTP?
More importantly, if RIM was going after all of these other companies, then it was hardly "novel", right?
Neither company deserves a patent in this case (which appears to be the case with about 98% of all software patents).
"and I wonder if IBM will market the cell (or a variant thereof) to the HPC market"
Products are already on the market from Mercury Computer Systems:
"Mercury's first Cell BE processor-based product is available for industrial, medical, and military markets. The Dual Cell-Based Blade offers outstanding performance for high-performance computing (HPC) applications. Performance scales dramatically when the application is distributed across multiple Dual Cell-Based Blades in IBM's industry-leading BladeCenter® platform or across the network."
Itanium has been taking market share from Power????
"Sales of IBM's Unix systems, called the pSeries, grew 15% in the first quarter and 36% in the second quarter--far outpacing Sun and HP. The trend should continue in the fourth quarter--historically, industrywide Unix sales have spiked 25% during this period--and into 2006, when IBM introduces a new high-end chip called Power5+."
Intel's biggest failure was taking so dang long to get the Itanium to market. I remember when it was annouced, and a short decade later the thing arrived. By the time it arrived things had changed quite a bit at HP, Intel and the computer markets in general.
"It is the closest thing we humans have created so far to Artificial Intelligence."
While I like Google too, I think you should educate yourself regarding AI research.
Niagra appears to be a good CPU. However, it is designed with a different goal than the Cell. Niagra has 8 fully functional cores operating at lower speeds and designed with multiple-connection servers in mind (web servers). The Cell processor has 1 fully functional CPU with 8 support processors, all of this designed for high-speed processing and high-bandwidth memory transfer (think video processing).
Most likely, you will not see a Niagra processor in a gaming unit, and you won't see a Cell CPU in a business server, those are not the functions they were designed for.
A little bit of evidence that the Cell will probably work as advertised is that Mercury Computer Systems is shipping Cell based evaluation systems intended for the following industries:
"medical imaging, industrial inspection, aerospace and defense, seismic processing, and telecommunications"
Given that companies (both sellers and buyers) perform due diligence before jumping into something like this, and IBM's track record with processors (especially recently, Power4, Power5, etc.), I'm inclined to believe Cell won't disappoint.
The "randomness" could very well be introduced by normal physical means (such as radiation from Sun altering molecules within cell, which we know does happen). The use of the word "random" merely means that the causative effect is so minute and detailed that without tracking every single particle in the Universe, we can think of it as "random" from our macro point of view.
Now there is the issue of randomness within quantum mechanics, but I assume you are also arguing that our understanding of QM is wildly incorrect also.
Forgive me, but I am just stunned by these types of questions.
If I said, "provide me an example of a situation where a GUI would be quicker/easier/fewer steps/less error prone/whatever compared to a command line interface, and I will give you $100", you don't think you could come up with even 1?
I personally could fill a book of pro's for both GUI and command line.
When IBM finally adds rollback segment to DB2 then I won't be as upset about DB2 being pushed onto my real time projects
You seem to be implying that Oracle's implementation of it's roll back mechanism via rollback segments is a good thing?????
Don't get me wrong, I like Oracle's concurrency model, but the implementation using roll back segments is far too problematic to consider it a good thing.
Intel Hyperthreading is a response to their pipeline problems. It is not simultaneous multi-threading, it is merely caching registers and switching quickly. 15% is about right for that approach.
IBM's SMT (and possibly Sun's, not sure) is an actual duplication of much of the processing logic and allows for better performance gains.
But the PowerPC core does have a more limited pipeline and branch prediction logic than some of the other Power chips. I believe this simplification was to make room for other "stuff."
Re:Since the submitter didn't bother to explain...
on
IBM Releases Cell SDK
·
· Score: 1
Not "useless otherwise."
I am working on a project that is cpu bound with mostly non-"matrix-vector" math, but is highly "parallelizable" (I know, it's probably not a word). I'm looking forward to the raw speed (of instructions and memory xfer) and multi-processing from the cell.
PowerPC instruction set is a subset of Power
AIX runs on Power, not PowerPC
How can AIX trap an instruction that is already supported by it's processor?
About Power Architecture
Power Architecture(TM) technology is an instruction-set architecture that spans applications from consumer electronics to supercomputers. Power Architecture encompasses PowerPC®, POWER4(TM) and POWER5(TM) processors.
What makes this microprocessing technology so different -- and such a breakthrough? Scalability, flexibility and customization. Power is built on an open architecture, making it an open ecosystem that supports freedom of action in the marketplace.
Look at the sheer amount of money Intel and HP have dumped into IA64, Sure, they can get some impressive results, but I suspect that given THAT amount of cash, time, and engineering, x86, Alpha, Power, etc all could have reached at least the same performance level.
I agree with your post but would also point out that Power5 already exceeds Itanium2 (significantly) (benchmarks TPC, SAP, etc.)
Clearly measuring stuff is important, but I would disagree with your statement. There is considerable "knowledge" that is not easily quantified.
The human brain is a complex neural network which operates by finding local minima in an N dimesional space, thereby matching inputs to best outputs for achieving desired goals based on prior experience.
Using that system (which is not really a system of quantification), humans can become "experts" and make very good decisions about the field they are focused in.
Clearly the same applies, to a degree, to the IT field. Not to say there can't be hidden costs or unknowns, but at the same time a person can choose the "best" option based on a combination of many variables measurable and not so measurable.
Slashdot Mirror
From post above: There are some decent ideas in the proposal: - Novelty as the only condition
. .me. really.)
This is hardly a good idea.
As a matter of fact, with respect to software, novelty is almost guaranteed in that everything you do is just a big set of combinations of the various building blocks. At some level, just about anything could be argued to be novel and has been.
For example, combining internet purchasing, clicking a button and the idea that fewer clicks is better, you end up with the novel (barely) method of 1 click purchasing. Hell, I might as well write a patent generator program that will spew out all combinations of various methods/sequences/actions, cherry pick the ones I think look most valuable and buy my lottery tickets by patenting all of this crap.
By eliminating the non-obvious requirement you have just made an incredibly flawed system (because there isn't enough staff to check up on non-obvious properly) into the worst of all possible worlds.
The only saving grace is that each patent will expire after 4 years.
These are the types of patents that must not be allowed under any patent system:
1) Amazon one-click (gee, who would have thought that fewer clicks would be better)
2) NTP and RIM for e-mail patents (transmitting data over an existing transmission medium???? Are you serious???)
3) System for keeping track of cell call minutes and stopping service when maximum reached
4) Method for distributing nutritional information over mobile devices (you...have...got...to...be...freaking...kidding.
"I'd never want the FULL version of these databases, even if those, too, were free"
Well, in an enterprise application environment you would want them. In an enterprise application environment, footprint is near the bottom of the list of priorities. The top of the list include:
Reliability Scalability Performance
In the 80's and 90's IBM service reps ran around with something they called a "brick" which was a wireless device that they used to communicate with the main office, wouldn't that be prior art for RIM and NTP?
More importantly, if RIM was going after all of these other companies, then it was hardly "novel", right?
Neither company deserves a patent in this case (which appears to be the case with about 98% of all software patents).
"and I wonder if IBM will market the cell (or a variant thereof) to the HPC market"
Products are already on the market from Mercury Computer Systems:
"Mercury's first Cell BE processor-based product is available for industrial, medical, and military markets. The Dual Cell-Based Blade offers outstanding performance for high-performance computing (HPC) applications. Performance scales dramatically when the application is distributed across multiple Dual Cell-Based Blades in IBM's industry-leading BladeCenter® platform or across the network."
Itanium has been taking market share from Power????
"Sales of IBM's Unix systems, called the pSeries, grew 15% in the first quarter and 36% in the second quarter--far outpacing Sun and HP. The trend should continue in the fourth quarter--historically, industrywide Unix sales have spiked 25% during this period--and into 2006, when IBM introduces a new high-end chip called Power5+."
The designers of the AS400 didn't go to 2048 bits, but every address is 128 bits from day 1 for just the reason you mention.
Intel's biggest failure was taking so dang long to get the Itanium to market. I remember when it was annouced, and a short decade later the thing arrived. By the time it arrived things had changed quite a bit at HP, Intel and the computer markets in general.
"It is the closest thing we humans have created so far to Artificial Intelligence." While I like Google too, I think you should educate yourself regarding AI research.
Niagra appears to be a good CPU. However, it is designed with a different goal than the Cell. Niagra has 8 fully functional cores operating at lower speeds and designed with multiple-connection servers in mind (web servers). The Cell processor has 1 fully functional CPU with 8 support processors, all of this designed for high-speed processing and high-bandwidth memory transfer (think video processing).
Most likely, you will not see a Niagra processor in a gaming unit, and you won't see a Cell CPU in a business server, those are not the functions they were designed for.
A little bit of evidence that the Cell will probably work as advertised is that Mercury Computer Systems is shipping Cell based evaluation systems intended for the following industries:
"medical imaging, industrial inspection, aerospace and defense, seismic processing, and telecommunications"
Given that companies (both sellers and buyers) perform due diligence before jumping into something like this, and IBM's track record with processors (especially recently, Power4, Power5, etc.), I'm inclined to believe Cell won't disappoint.
The Xbox and PS3 are not the same machine.
The Cell and the Xenon (or whatever it's called in the Xbox360) take different approaches for game optimization in their internal workings of the CPU.
So each of the quantum computer advances that have been announced over the last 8 years by IBM, HP and others were all stunts also?
Do scientists not routinely publish information gained from doing research?
Is this not exactly how scientists are able to build on each others work to rapidly move forward in difficult fields like this?
I'm stunned at the ignorance.
You seem hung up on the term "random".
The "randomness" could very well be introduced by normal physical means (such as radiation from Sun altering molecules within cell, which we know does happen). The use of the word "random" merely means that the causative effect is so minute and detailed that without tracking every single particle in the Universe, we can think of it as "random" from our macro point of view.
Now there is the issue of randomness within quantum mechanics, but I assume you are also arguing that our understanding of QM is wildly incorrect also.
Forgive me, but I am just stunned by these types of questions.
If I said, "provide me an example of a situation where a GUI would be quicker/easier/fewer steps/less error prone/whatever compared to a command line interface, and I will give you $100", you don't think you could come up with even 1?
I personally could fill a book of pro's for both GUI and command line.
A GUI and a command line interface are both very important, but for different reasons and different situations.
If that was really just some sort of a joke, I apologize for not getting it.
PowerPC's ancestor is POWER from IBM, not the 88000 from Motorola. Motorola's part of the AIM alliance was manufacturing, not design.
When IBM finally adds rollback segment to DB2 then I won't be as upset about DB2 being pushed onto my real time projects You seem to be implying that Oracle's implementation of it's roll back mechanism via rollback segments is a good thing????? Don't get me wrong, I like Oracle's concurrency model, but the implementation using roll back segments is far too problematic to consider it a good thing.
Intel Hyperthreading is a response to their pipeline problems. It is not simultaneous multi-threading, it is merely caching registers and switching quickly. 15% is about right for that approach.
IBM's SMT (and possibly Sun's, not sure) is an actual duplication of much of the processing logic and allows for better performance gains.
Not all vendor's solution is the same.
But the PowerPC core does have a more limited pipeline and branch prediction logic than some of the other Power chips. I believe this simplification was to make room for other "stuff."
Not "useless otherwise."
I am working on a project that is cpu bound with mostly non-"matrix-vector" math, but is highly "parallelizable" (I know, it's probably not a word). I'm looking forward to the raw speed (of instructions and memory xfer) and multi-processing from the cell.
Just wanted to point out that not all software that runs on Linux is open source.
Why is there an attempt to force drivers to be open source when I can write an ERP package that runs on Linux that isn't open source?
PowerPC instruction set is a subset of Power
AIX runs on Power, not PowerPC
How can AIX trap an instruction that is already supported by it's processor?
From the IBM website:
About Power Architecture
Power Architecture(TM) technology is an instruction-set architecture that spans applications from consumer electronics to supercomputers. Power Architecture encompasses PowerPC®, POWER4(TM) and POWER5(TM) processors. What makes this microprocessing technology so different -- and such a breakthrough? Scalability, flexibility and customization. Power is built on an open architecture, making it an open ecosystem that supports freedom of action in the marketplace.
Look at the sheer amount of money Intel and HP have dumped into IA64, Sure, they can get some impressive results, but I suspect that given THAT amount of cash, time, and engineering, x86, Alpha, Power, etc all could have reached at least the same performance level.
I agree with your post but would also point out that Power5 already exceeds Itanium2 (significantly) (benchmarks TPC, SAP, etc.)
Clearly measuring stuff is important, but I would disagree with your statement. There is considerable "knowledge" that is not easily quantified.
The human brain is a complex neural network which operates by finding local minima in an N dimesional space, thereby matching inputs to best outputs for achieving desired goals based on prior experience.
Using that system (which is not really a system of quantification), humans can become "experts" and make very good decisions about the field they are focused in.
Clearly the same applies, to a degree, to the IT field. Not to say there can't be hidden costs or unknowns, but at the same time a person can choose the "best" option based on a combination of many variables measurable and not so measurable.