Intel and Dell presented servers using the “Tulsa” Xeon processor. The new chip has been used in Intel’s server and a Dell PowerEdge 6850 on display at the LinuxWorld Conference and Expo. The Dual-core processor Tulsa is the last of the ill-fated NetBurst lineage of x86 chips from Intel.
Intel has begun selling Tulsa this summer. Each Tulsa processor core comes with 1MB of level-two cache memory, and the 2 cores share a whopping 16MB of level-three cache memory, more than any other x86 processor. Cache memory stores instructions and data. This way the information can be retrieved more quickly.
According to Intel, Tulsa uses 1.3 billion transistors, nearly as many as the 1.6 billion in Intel’s new “Montecito” Itanium processor. News.com technology news web site explains that Intel can afford to build caches larger than those made by its rivals, because it has moved more quickly to a new manufacturing process with 65-nanometer circuitry elements, letting more transistors be crammed into a given surface area. Most of the Intel’s rivals still are building chips with a 90-nanometer process.
According to Intel, the Tulsa runs at a top speed of 3.4GHz. It comes in two models – a high-performance one that consumes 150 watts of power and another geared for rack-mounted servers that consumes 95 watts.
Compared with Paxville, Tulsa boosts transaction processing performance by a factor of 1.7, enterprise resource planning software by a factor of 1.4, and e-commerce software by a factor of 2.
AMD Is Looking For For Quad-Core Processor
The Intel’s rival AMD announced the first major upgrade of Opteron server processor. It includes hardware-assisted virtualization and use of second-generation double data rate memory.
The company believes that Opteron Rev-F is an enhanced processor platform that can compete Intel Woodcrest sever chip. AMD’s new Opteron receives positive reponse from server makers Hewlett-Packard, IBM, and Sun Microsystems.
The the analytical agency Mercury Research says that AMD currently has a 26% share of the x86 server processor market, up from 22% in the beginning of this year. The agency believes that AMD would increase that share until the end of the year.
AMD has also completed the design of its quad-core Opteron processor. The new line is scheduled for introduction in 2007. AMD says it will be able to drop in the new AMD Socket F with virtually no changes to their motherboards. According to the company, some application may experience as much as a 20% performance gain with the new Opteron Rev-F processors compared to the existing Opterons. AMd however expects a boost when the quad-core chip becomes available.
The Socket F is the first socket transition for Opteron, introduced in 2003. Opteron Rev-F maintains the same architecture as Opteron line unlike Intel’s new Core-based processor, which represents a break from Intel’s older NetBurst architecture.
The Opteron Rev-F will be AMD’s first processor to use DDR2 memory, which necessitated the socket change. Intel is using DDR2, but it is now transitioning to fully buffered DIMMs, or dual in-line memory modules.
According to Server Pipeline The Opteron Rev-F will also be AMD’s first processor to include embedded or hardware-assisted virtualization. The technology, which was developed under the code name Pacifica, is similar in approach to the embedded virtualization technology Intel has included in its latest-generation processors.