Advanced Micro Devices, the world’s second largest producer of x86 microprocessors, on Thursday has demonstrated its first quad-core microprocessor for servers and high-end workstations. The chip features the so-called native quad-core design and is being shown off inline with AMD’s promises, yet, later than the company announced originally.
AMD Quietly Demos Quad-Core Chips
The code-name of the new processor, according to media reports, is Barcelona and it is designed for multi-processor servers, which means that a four-socket machine now has the power of sixteen processing engines, an amount that only very high-end machines featured just two-three years ago.
So far there is no much information about the new chip. It is known, however, that the new microprocessor is made using 65nm process technology and is drop-in compatible with mainboards and servers featuring socket F (LGA1207. The chip was reportedly showcased at a meeting with various analysts, but details like core-clock and other peculiarities are currently uncertain.
The demonstration of the quad-core chip from AMD comes as the company’s Quad FX platform failed to meet expectations of the enthusiast community.
AMD’s Quad-Core: Through Thorns to the Stars
AMD is behind its own track in case of quad-core microprocessors. Earlier this year the company planned to demonstrate its quad-core chips on the socket F platforms during the launch of dual-core AMD Opteron processors with DDR2 memory support.
Earlier it was reported that the first quad-core chip from AMD is code-named Deerhound, which will be intended for socket F infrastructure and will have shared level-two cache along with dual-channel registered DDR2 memory controller. The new chip is expected to use the new code-named K8L design, which features some improvements.
According to AMD, K8L includes a quad-core design for servers, workstations and high-end desktops, and a dual-core design intended for mainstream desktop markets. These next-generation processors will be built using AMD’s 65nm Silicon-on-Insulator (SOI) fabrication process, and include a broad range of functionality and micro-architectural improvements, including a new ability to dynamically alter the frequency of each core on the chip to match application workloads and thereby reduce overall power consumption.
AMD believes that Intel’s approach to put two dice on a single slice of substrate to build a quad-core processor - like Intel Corp. did with its Intel Core 2 Quad/Extreme and Intel Xeon 5300-series processors is inefficient and AMD’s “native quad-core” design will provide better performance and scalability for servers.
Comments currently:
5
Discussion started: 11/30/06 05:06:30 PM
Latest comment: 12/31/06 05:11:19 AM
Expand all threads |
Collapse all threads
[1-3]
1.
Who would guess that AMD, who didn't even come out with the chip when they said they would and then all of a sudden the name has changed and functionality is "the same", would be bashing Intel saying that they didn't make a very efficient chip. The last time I checked Intel is the mother of the Microprcessor and has yet to be outdone.
[
Posted by: Stephen

|
Date: 11/30/06 05:06:30 PM]
+ expand thread (1 answer)
- collapse thread
Well just because you are the McDonald of the world doesn’t mean there aren’t better food else where.
Or Ferrari is bad vs GM just because GM sells more cars?
And Intel isn’t the mother of Microprocessors, IBM is.
[
Posted by: Kaz

|
Date: 12/06/06 02:09:33 AM]
2.
Why would they be "quiet" about it ***IF***
It was faster than Intel's solution?
It was of comparable performance for fewer watts?
Questions, questions. Perhaps their "native" quad isn't as good as Intel's two-die "inelegant" solution.
[
Posted by: Mark1

|
Date: 12/04/06 02:53:13 PM]
+ expand thread (1 answer)
- collapse thread
Maybe isn’t as good from performance point. But what you all Intel fans fail to see is that:
- 4 cores on the same die is actually technology advance, Intel solution is just a packing trick.
- Lower power consuming.
- More efficient.
- Better IO utilization, however no one has done real tests to see where and when the IO gets saturated on the Intel or AMD system.
Just look at what Intel says about their own quad systems:
- Each die is a dual core unit.
(doesn’t look better call Intel "quad", Double Dual Core)
- Faster time to market.
(Not really at least if we would do some comparing vs dual where AMD beat them)
- Less Engineering resources.
(Less or none, because its just a packing trick, there are no need of CPU designers here)
- Two smaller die yield better than a double sized die .
(Like I said should be called double core not quad, and by the way isn’t all this sounding very good just for Intel it self, I didn’t see me win anything here)
- Share wafer starts with dual core
(Good for me, wait Intel again)
- Ability to match die for better bin splits
(This is good for us, wait maybe not)
http://www.anandtech.com/showdoc.aspx?i=2840&p=4
One good thing that Intel “quad” could do would be “quad” at almost the same prices of dual like they did with the Pentium D, … but this time they are charging just 1,599$ for one (two double core) at New Egg.
http://www.newegg.com/Product/Product.asp?Item=N82E1681911 5011
[
Posted by: Kaz

|
Date: 12/06/06 02:32:47 AM]
3.
Optical-based processors will be the world's best processor in the near-future. Using light-technology, bottle-necking can be greatly reduced. Now IBM and Intel are doing this research separately. Hope both can come out with one of the best ever.
[
Posted by: Tangent

|
Date: 12/31/06 05:11:18 AM]
[1-3]