The chemical processes are fairly slow, but the human bran is massively parallel and adaptable.
The human brain processes huge amounts of information and sends countless instructions throughout the body every second, all while running the most advanced AI ever known. It also learns both through said AI's concious effort, as well as automatically. As far as we can tell, it even has effectively infinite storage capacity.
The way computers process data is entirely alien to the human brain. Numbers are an abstract concept, which our brains must "emulate". Numbers and math are also generally serial in nature, which does against how our brains function. So, it's no wonder that a human brain isn't as good with numbers as a computer specifically built to handle them.
The Inquirer.com isn't exactly a bastion of responsible reporting.
It doesn't look like an interview took place at all. It looks like they took some choice quotes out of context from the kernel development mailing list to spur some pageviews.
If the average gate uses 4 transistors (let's say 8 to be conservative), then the P4 Northwood's 55 million would mean the approx. $300 P4 chip would cost around $10,000 if made out of those fpga elements. I'd hardly call over 30x the cost per gate economical.
FPGA prices still have a long way to fall yet before they're more economical than simply getting a fast ASIC and emulating your target enivornment.
To look at things from another perspective-
I've yet to be given a rigorous proof of why intellectual property is neccessary in the first place. I've heard several arguments, but never a rigorous proof.
The problem with the ISO 9002 standard is that it only requires that your quality assurance is consistent. You could be consistently bad and qualify for ISO 9002 certification.
Slashdot Mirror
The chemical processes are fairly slow, but the human bran is massively parallel and adaptable.
The human brain processes huge amounts of information and sends countless instructions throughout the body every second, all while running the most advanced AI ever known. It also learns both through said AI's concious effort, as well as automatically. As far as we can tell, it even has effectively infinite storage capacity.
The way computers process data is entirely alien to the human brain. Numbers are an abstract concept, which our brains must "emulate". Numbers and math are also generally serial in nature, which does against how our brains function. So, it's no wonder that a human brain isn't as good with numbers as a computer specifically built to handle them.
FYI: They're paid to run the site. It's their job.
The Inquirer.com isn't exactly a bastion of responsible reporting.
It doesn't look like an interview took place at all. It looks like they took some choice quotes out of context from the kernel development mailing list to spur some pageviews.
If the average gate uses 4 transistors (let's say 8 to be conservative), then the P4 Northwood's 55 million would mean the approx. $300 P4 chip would cost around $10,000 if made out of those fpga elements. I'd hardly call over 30x the cost per gate economical. FPGA prices still have a long way to fall yet before they're more economical than simply getting a fast ASIC and emulating your target enivornment.
This is a dupe.
I remember people mentioning that this thing is uneconomical compared to an emulator because of how expensive FPGAs are.
To look at things from another perspective- I've yet to be given a rigorous proof of why intellectual property is neccessary in the first place. I've heard several arguments, but never a rigorous proof.
The problem with the ISO 9002 standard is that it only requires that your quality assurance is consistent. You could be consistently bad and qualify for ISO 9002 certification.