AMD Dominates
Shortening Data Paths
This cut down considerably on latency the time it takes to arbitrate data priorities and routing. AMD also upped the amount of L1 cache and, in the FX57, shrank the size of the processor die from 130nm to 90nm. Even if none of the other design improvements helped to speed data throughput in the FX (and they all did), just the downsizing of the die would have done it.
Think of it in terms of walking from Point A to Point B. Suppose it takes you an hour to do that. Then the distance between A and B is shortened by 30 percent. Now it only takes you 40 minutes to make that same trip, and you won’t break a sweat because you haven’t increased your walking speed at all. In terms of a CPU, that means an immediate performance boost without needing to raise the CPU’s clock speed and that equates to more performance without additional heat.
Punted Into Dual Core
For the FX60, however, dual core means a compromise. While the older FX57 is clocked at 2.8GHz, the FX60 has been backed down to 2.6GHz. Technically, that means the FX60 should be slower and it is, when gaming. The truth of the matter, however, is that is the type of “slower” that’s the result of counting benchmark numbers. If it’s noticeable at all in real life, the operative word would be “hardly.”
In exchange, AMD’s move to dual core has let it retain its title as a great gaming processor and also allowed it to take on Intel’s dual-core desktop audience in the productivity arena. It might even be apropos to call the FX60 the world’s first Swiss Army processor. The only trick is building a system around it.
First, though, a bit of background. AMD has dominated the gaming market ever since its Athlon processors first appeared. Whether it’s because of some inherent superiority in the chip that gives it an advantage over Intel’s Pentium 4, or even if it's just some clever trick of program compilation that uses an optimization that favors the Athlon and not the Pentium, this means that Athlon is definitely the frontrunner. And the FX line has always been the crown prince of Athlons.
What AMD did with the FX series (FX51, FX53, FX55, and FX57) was to develop a process of getting data into and out of the CPU as quickly as possible. It moved the memory controller and the HyperTransport bus controller inside the chip, and then added a construct called a crossbar switch so that the majority of the arbitration needed to send data where it needs to go is done inside the processor at CPU speeds, not outside at slower bus speeds.
According to everyone in the know, the next processor on the drawing board was supposed to be a single-core FX59. That would have made sense, given the numbering progression thus far. Why did AMD jump to a dual-core FX60? Probably one of the main reasons was because it could. Figuratively speaking, the Athlons, including the FX versions, have always had a blank spot inside that was just waiting for a second core. It was just a matter of “writing” it into the die. That’s how AMD got the jump on Intel in the dual-core arena in the first place.
Page 2:
Creating A Nemesis
![]()
1
|
2
|
3
Next Page »
Stay connected and informed by visiting the CA Solutions Center Community!

Become a member today for instant access to free InformationWeek research, expert advice, peer perspectives, and more on the following topics:
- Application Performance Management (APM)
- Security Management
- Mainframe 2.0
- IT Automation
- Service Assurance
Also, visit our Government and Financial Services groups to see how these technologies apply specifically to those industries.
NOTE: Offer valid for U.S., U.S. possessions, & Canada only.