Quote:
|
Originally Posted by kooloser
No, the Joes make Intel money.
|
The retail market with the Best Buy's and the Circuit City's hovers around 50/50 Intel/AMD, with Intel on the north side of 50%.
|
Originally Posted by kooloser
Man, that just the history is confusing. The fact is AMD has 4 sockets (unless we count the MP), and will soon have 3. That is clear and concise. Whats even better is the normal users will only have to worry about 2. If the buyer is really computer illiterate, they'll only have to worry about 1. Hows that for simple.
Why didn't AMD stick with the 940? Obvious reasons, number one being that they didn't want it shared with the Opteron. Also, the ECC was too expensive and hard to find for normal people. That warrants a perfect reason to switch to 939. Of course they couldn't have launched 939 with the AMD 64s because of a clear shortage of parts, it was just easier to have them plug into the 940. |
|
Originally Posted by kooloser
No, the Joes make Intel money.
|
|
Originally Posted by Nychold
Another reason Socket 940 is worse than Socket 939 is a socket 940 motherboard requires a 6 layer process, where the socket 939 only requires 4 layers. That makes 939 motherboard cost less, and easier to design. All good things for desktop computers.
![]() |
|
Originally Posted by Superguy
Read this article talking about the takeup of the Itanium. It's a little old, but it shows that Itanium take up IS happening and was happening as early as last year
|

|
Originally Posted by Superguy
Bottom line: Itanium might sell less systems, but the ones they sell have a LOT more processors in them and scale a LOT better.
|
| The Dawning 4000A supercomputer is planned to run at maximum speeds in excess of 10 trillion operations per second, or 10Tflops. It is also expected to be the first supercomputer made in China to be considered among the most powerful in the world. Coupling the AMD Opteron processor’s high performance, 64-bit capabilities and powerful, highly scalable design advantages with Dawning’s strength in research and development, the Dawning 4000A project represents the first IT industry effort in China to build one of the world’s most powerful supercomputers. More than 2000 AMD Opteron 800-series processors are planned to power the system with planned memory capacity to be up to 2256GB, or 2.256 Terabytes. |
| The Sandia computer, called Red Storm, is being designed to help the National Nuclear Security Administration work out complex weapons problems. The Sandia system is expected to be one of the fastest computers ever made, able to perform about 40 trillion calculations per second (40 teraflops). Customers other than Sandia could get a system of different size, but still built with Red Storm's 64-bit processor, designed by Advanced Micro Devices Inc. and based on Linux. |

System Facts:
|
| A 100-teraflop (trillions of calculations per second) Cray system at Oak Ridge is planned for 2006, with the potential to grow to 250 teraflops in 2007. Near-term plans call for increasing the capacity of the current Cray X1(tm) supercomputer at ORNL to 20 teraflops in 2004, with a 20-teraflop Red Storm-based system from Cray added in 2005. The systems will be housed in ORNL's new National Leadership Computing Facility in Oak Ridge, Tennessee. |
| As one of the largest cluster supercomputers ever built, Lightning will consist of 2,816 Advanced Micro Devices Inc. Opteron processors and 1,408 dual-processor nodes, all of which will be interconnected by Myricom Inc.'s Myrinet high-speed network. The project, with a total value of nearly $10 million, will be the first 64-bit Linux supercomputer in the ASC program. Lightning is designed to have a theoretical peak speed of 11.26 trillion calculations per second. This speed will certainly rank Lightning among the top 10 supercomputers in the world and will give researchers at Los Alamos more computing strength, said Dean Hutchings, Linux Networx's chief operating officer. |

Cluster Overview:
|
| The “Orange” cluster will be part of Los Alamos’ Institutional Computing project that supports scientific, medical and environmental research such as the design of antibiotics and simulations of wildfires and water resources. Orange, a 256-node dual-processor cluster, is expected to be the first large-scale AMD Opteron processor-based cluster using InfiniBand technology for greater interconnect bandwidth and scalability. Both the “Lightning” and “Orange” clusters are being designed, built and integrated for Los Alamos National Laboratory by LinuxNetworx and will be powered by the AMD Opteron processor Model 244. Both clusters will utilize the Arima HDAMA mainboard. |
