Open-PC, designed by and for the Linux community
The project Open-PC has shown its prototype, which will be available from February.
The Open-PC initiative has created a PC designed especially for Linux users. Has been molded with suggestions and community surveys to reach the final hardware and software.
In the absence of a few weeks to start sales, calling on the community that promote the extent possible.
New Green500 list
He has published a new list of computer facilities with greater energy efficiency, the list Green500 .
IBM continues to dominate the top of this list, noting that the top six are for facilities based on PowerXCell 8i processors.
Seventh place for a team with the unusual architecture GRAPE-DR , and just after the first computer with x86_64 architecture, accelerated by ATI Radeon HD.
The following fifteen infrastructure solutions based on IBM Blue Gene / P.
Cray heads the new Top500 list
The new TOP500 list was published today. With it comes the surprise of IBM is not first, but has been Cray supercomputer Jaguar.
Jaguar has built for the computer science center of Oak Ridge. It is built on AMD 6-core Istanbul , for a total of 1,759,000 GFlops (224 162 nuclei).
Other headlines :
402 systems use Intel processors.
427 are carried x86 quad-core processors.
IBM is a leader in installed capacity, 35%, followed by HP, Cray and SGI.
Micros with a hundred nuclei
The Californian company level up its line of high performance microprocessors increasing their power while reducing energy consumption. The third-generation microprocessors Tilera TILE-Gx, and come after Tile64 Tile32Pro and can incorporate 16, 36, 64 and 100 centers offering energy efficiency "ten times the Westmere", the next generation of Intel processors. They will be manufactured by the leading global foundry, TSMC, under 40-nanometer process technology and its commercialization is expected in late 2010.
Tilera executives say they are
"Four years ahead of Intel in microprocessor design," because they have been able to start from scratch as opposed to Intel "has a huge investment in trying to ensure x86 compatibility for servers and micro earlier sector consumption. "
Read more
The concept behind Barrelfish
As OS design seems to have been understood in some blogs that are just as multicore or SMP system.
The concept behind the other Barrelfish distributed systems coming from Microsoft Research is an interesting technology news: a multikernel able to communicate with user space and with x86 processors, ARM, CELL or graphics chips and all constitute a single system eyes of those who operate it.
Then this system in a primitive state seeks unprecedented modularity in distributed: it is not an implementation advantage quadcores resources, it is spread in heterogeneous clusters or configuration NUMA (non-parallelizable) by intercom protocols URPC of a group of nuclei coupled in one general system.
Is being planned under a BSD license (open but sealable).
The only source that explains in detail ( in English )
Does anyone remember the decentralized modular system try without kernel and 100% rechargeable without restarts and never moved from theory? Unununium called and returned (assembler and BSD license). Though still but slightly profiled are the initial self-replicating systems (see the monolith that goes "hyper" when you get Linux 2019).
Is the universe a colossal Turing machine?
A MT is a formal model of the concept of algorithm we can imagine how a device can read / write symbols from / to a tape, and move through it (or cause the tape to move, which gives it a thing as other).
The tape can imagine the unlimited length, and divided into squares, each of which may be a symbol.
The MT is at all times in a certain state
internal (from a certain set of possible states) and has an internal transition function that tells you what to write, how to move, and how to change the internal state based on the current state and what they are reading in each time of the tape.
The definition of TM is incomplete if we indicate the alphabet A that contains all the symbols that can be read / write, the set of possible states Q, the
q0 initial state and the transition function d.
The operation of the MT terminates when it reaches a stop state.
The initial contents of the tape corresponding to the input of the algorithm, and the contents to its output end.
Decoding the Universe, Charles Seife (for whom it was anime)
Programming the Universe, Seth Lloyd (for those who do not get discouraged)
Source: Microserfs
GenodeOS 2009.09 liveCD
There is already a second image of this prototypical operating system from the previous version in October 2008.
We are invited to test it with KVM from your front-end qemu among other options would be to invest in Linux paravirtualizarlo purpose, but given their immaturity is the resource that calls for the moment.
The purpose of Genode Nitpicker is mounted on OKL4 microkernel, Fiasco and / or Pistachio can run under GNU / Linux, which we refer to this as trite as failed religious purpose of the truncated Stallman Do not see the connection between Genode Nitpicker and Hurd -Mach?
Still no link with any portable devices and systems needed to flush.
That is why some hackers go injecting Ubuntu Linux hardware and the Amazon Kindle 2 for storing and managing e-books platform.
Platform expansion
Filed under: Hardware , Linux , Software , Operating Systems
Opera 10 is ready for those who prefer its interface, improvements in javascript from 9.5 Elder Futhark (the alphabet of the proto-Norse) or function to read in Opera Presto 2.2 - 100% Acid 3 - tags and statements in the language of the web, as do Safari, Firefox, IE8 or Chrome. The latest stable version was 9.64.
+ GPU parallelizing CPU load and process
It is known that any general purpose CPU to execute instructions and any GPU specific code only works with their units or cores of stream flow. Well AMD has lead to software that is coming forward in hardware: x86 OpenCL SDK. Still in beta phase involves the transport to other systems that only OpenCL in Snow Leopard could do. And as the computationally intensive work is split between dual / quad-core Athlon, Phenom, Core2, Core i5-i7 (and upcoming 32nm i3) with integrated chipsets AMD 790GX IGP type / nForce 980a and / or GeForce graphics and Radeon, NVidia Ion well, AMD's Fusion and Pine Trail Atom + GPU will share the work harmoniously and synchronously to development on this tool in games, 3D graphics editors, HD video decoding, integer factorization and others ... juicers. The external factor to this is, that all this new hardware now in Argentina will be a minimum 30% more expensive for a new tax system that aims to develop internal and bury Alvin Toffler .
MorphOS 2.3 & Power7
Although the new PowerPC-and not accumulated as just five years ago, platforms Pegasos I G3 (750CXe IBM), Pegasos II (Freescale MPC7447) and Efika (Freescale MPC5200B) with such Motorola processors can test the latest build of MorphOS 2.3 , which provides some of its features here and there . Join now Genesi hardware using Cortex A8 ARM SoC from Freescale i.MX515 chip.
It is estimated that there will be availability of IBM Power7 and 4.4 GHz in 45nm mid-2010. Consist of 2, 4 or 8 cores with 4 threads each resulting in a peak computing power of 258.6 GFLOPS per chip, much more than they demand a meeting of MorphOS.
Tira Ecol
Translator
