Dirtynap
09-26-2006, 07:40 PM
The below is pure hardware news:
Intel today formally annouced the release of Quad core processing in November 2006.
The first of these processors will be the Kentsfield core named Core 2 Quad not Quadro as some have been rumouring, and will be under the extreme brand. The processor is believed to run at 2.67GHz (basically the chip is two E6700) although some believe on release the chip will be clocked at 2.4GHz. Overclocking on test chips has already seen people hit 4Ghz on each core (http://www.xtremesystems.org/forums/showthread.php?t=110121). However it is believed the chips can go even further especially when people have pushed X6800's and E6600 to beyond 5GHz (http://www.overclocking-masters.com/index.php).
Along with the Kentsfield there will be Clovertown, a quad core Xeon processor and Tigerton a quad core Xeon processor designed for servers with more than two chips. More on the Xeon chips can found here (http://news.zdnet.com/2100-9584_22-6038148.html).
Besides the new quad-core CPUs. Intel also disclosed that their upcoming 45-nm manufacturing process is on track for production in the second half of 2007. Intel CEO Paul Otellini even went as far as to disclose that the company has 15 45-nm products already in development in the desktop, mobile, and enterprise segments. The first of these CPUs is "on track to complete its design in the fourth quarter of this year".
Going yet further, Otellini also provided a glimpse into their next-generation CPU micro-architecture beyond today's Core 2 CPU, coming in 2008 is their 45-nm "Nehalem" CPU, which will be followed up in 2010 by "Gesher".
To demonstrate how Moore’s Law (http://en.wikipedia.org/wiki/Moore) will continue well into the future with amazing potential, Otellini showed a new research prototype processor that has 80 floating point cores on a single die, each running at 3.16GHz. The tiny silicon die is just 300mm², and is capable of achieving a Teraflop, or 1 trillion floating point operations per second. He contrasted this with Intel’s historic breakthrough 11 years ago with the world’s first Teraflop supercomputer, a massive machine powered by nearly 10,000 Pentium Pro processors in more than 85 large cabinets occupying about 2,000 square feet. Todays fastest stock processor the X6800 Core 2 Extreme can do around 18,712 MFLOPS (http://72.14.221.104/search?q=cache:gFHxU8F47TYJ:www.computerpoweruser. com/articles/archive/c0610/25c10/25c10chart.pdf%3Fguid%3D+core+2+duo+X6800+MFlops&hl=en&ct=clnk&cd=1) (which is 18,712 million floating point calculations per second, or 18.7 american billion roughly, which is 18.7 GFLOPS). So the little thing is mighty mighty fast.
So the future is bright for processing speeds at least.
Read more here (http://www.intel.com/pressroom/archive/releases/20060926corp.htm).
Additional Source (http://www.firingsquad.com/news/newsarticle.asp?searchid=12432)
Additional Source includes images and information about laser FSB's (http://www.anandtech.com/tradeshows/showdoc.aspx?i=2841)
For those wondering what a laser FSB is, its exactly what it says. A laser is generated on the silicon die and used to encode binary data this is then fired at a decoder to turn it back into electrical signals for use by the processor itself. Think of the laser FSB much like an optic fibre and then think of how much bandwidth you can get down an optic fibre. The technology is years and years away yet but the potential is enormous.
Intel today formally annouced the release of Quad core processing in November 2006.
The first of these processors will be the Kentsfield core named Core 2 Quad not Quadro as some have been rumouring, and will be under the extreme brand. The processor is believed to run at 2.67GHz (basically the chip is two E6700) although some believe on release the chip will be clocked at 2.4GHz. Overclocking on test chips has already seen people hit 4Ghz on each core (http://www.xtremesystems.org/forums/showthread.php?t=110121). However it is believed the chips can go even further especially when people have pushed X6800's and E6600 to beyond 5GHz (http://www.overclocking-masters.com/index.php).
Along with the Kentsfield there will be Clovertown, a quad core Xeon processor and Tigerton a quad core Xeon processor designed for servers with more than two chips. More on the Xeon chips can found here (http://news.zdnet.com/2100-9584_22-6038148.html).
Besides the new quad-core CPUs. Intel also disclosed that their upcoming 45-nm manufacturing process is on track for production in the second half of 2007. Intel CEO Paul Otellini even went as far as to disclose that the company has 15 45-nm products already in development in the desktop, mobile, and enterprise segments. The first of these CPUs is "on track to complete its design in the fourth quarter of this year".
Going yet further, Otellini also provided a glimpse into their next-generation CPU micro-architecture beyond today's Core 2 CPU, coming in 2008 is their 45-nm "Nehalem" CPU, which will be followed up in 2010 by "Gesher".
To demonstrate how Moore’s Law (http://en.wikipedia.org/wiki/Moore) will continue well into the future with amazing potential, Otellini showed a new research prototype processor that has 80 floating point cores on a single die, each running at 3.16GHz. The tiny silicon die is just 300mm², and is capable of achieving a Teraflop, or 1 trillion floating point operations per second. He contrasted this with Intel’s historic breakthrough 11 years ago with the world’s first Teraflop supercomputer, a massive machine powered by nearly 10,000 Pentium Pro processors in more than 85 large cabinets occupying about 2,000 square feet. Todays fastest stock processor the X6800 Core 2 Extreme can do around 18,712 MFLOPS (http://72.14.221.104/search?q=cache:gFHxU8F47TYJ:www.computerpoweruser. com/articles/archive/c0610/25c10/25c10chart.pdf%3Fguid%3D+core+2+duo+X6800+MFlops&hl=en&ct=clnk&cd=1) (which is 18,712 million floating point calculations per second, or 18.7 american billion roughly, which is 18.7 GFLOPS). So the little thing is mighty mighty fast.
So the future is bright for processing speeds at least.
Read more here (http://www.intel.com/pressroom/archive/releases/20060926corp.htm).
Additional Source (http://www.firingsquad.com/news/newsarticle.asp?searchid=12432)
Additional Source includes images and information about laser FSB's (http://www.anandtech.com/tradeshows/showdoc.aspx?i=2841)
For those wondering what a laser FSB is, its exactly what it says. A laser is generated on the silicon die and used to encode binary data this is then fired at a decoder to turn it back into electrical signals for use by the processor itself. Think of the laser FSB much like an optic fibre and then think of how much bandwidth you can get down an optic fibre. The technology is years and years away yet but the potential is enormous.