Feeds

AMD claims no premium for four-way chips

The Opteron 6100 sales pitch

Top three mobile application threats

Here's a riddle for you: If you put two whole processors in a single socket, snuggling up tightly in a single ceramic package, and then if you glue four sockets together, is that a four-processor machine or an eight-processor machine? If you glue two sockets together using the same chips, is that a two-processor or a four-processor machine?

Yeah, that's what I was thinking, too. Maybe the Opteron 6100s are not as different from the Opteron 8400s as we might be thinking. And as long as software companies think of the Opteron 6100 4P as four processors instead of eight, that's what really matters to IT shops.

What also matters, of course, is how the Opteron 6100 platform stacks up against the Xeon 5600 and Xeon 7500 platforms. With the Xeon 7500s not yet in the field, AMD has to compare itself to the Xeon 5600s.

John Fruehe, director of product marketing for server and workstation products at AMD, says that the company is going to eliminate the "4P tax," by which Fruehe means that incremental cost in both the prior Opteron and Xeon product lines and the current Xeon 5600 line to move from 2P boxes to 4P boxes is gone.

The 2P workhorse machines can scale up to the 4P drafthorse machines, with the same chipset and literally the same processors, and most importantly, with the same processor prices. In the past, processors designated for four-socket boxes carried a pretty hefty premium. As the Intel Xeon 7500s most likely will unless the AMD Opteron 6100s force Intel to chop prices.

Looking at just two-socket comparisons between the Opteron 6100s and the Xeon 5600s, the AMD setup has eight memory channels (four per socket) and 24 DDR3 memory slots (three per channel), while the Xeon 5600 setup has six memory channels (three per socket) with 18 memory slots (three per channel again). The Opteron 6100 machine can be doubled up to four sockets, but the Xeon 5600 cannot.

Moreover, at current memory pricing, setting up a two-socket box with 96 GB of total main memory is about 34 per cent cheaper because AMD can do it with 4 GB DIMMs, while the Xeon 5600 machine has to use 8 GB DIMMs. (4 GB memory sticks cost under $200, and 8 GB sticks cost a little more than four times that at current street prices.)

That's the platform argument that AMD is making. Now, let's talk about pricing. Here is how AMD is stacking up the Opteron 4100s and 6100s against its prior generation of six-core Opterons and Intel's current Xeons, including AMD's expectations for the Xeon 7500s:

AMD Opteron vs Intel Xeon Pricing

As you can see, AMD is tipping its hand a little with the Opteron 4100s, and is planning on charging from $99 to $455 for these four- and six-core Lisbon processors, undercutting the current prices on its Opteron 1000 series and moving into the middle of the Opteron 2000 series. The Opteron 6100s start in the middle of the Opteron 2000s and extend about halfway up to where the Opteron 8000s used to be.

This is a big price cut for the chips, which will result in somewhat lower system prices. The pricing looks particularly aggressive compared to the current and expected prices for the complete Xeon lineup from Intel.

Seven Steps to Software Security

Whitepapers

Top three mobile application threats
Prevent sensitive data leakage over insecure channels or stolen mobile devices.
Implementing global e-invoicing with guaranteed legal certainty
Explaining the role local tax compliance plays in successful supply chain management and e-business and how leading global brands are addressing this.
Boost IT visibility and business value
How building a great service catalog relieves pressure points and demonstrates the value of IT service management.
Designing a Defense for Mobile Applications
Learn about the various considerations for defending mobile applications - from the application architecture itself to the myriad testing technologies.
Build a business case: developing custom apps
Learn how to maximize the value of custom applications by accelerating and simplifying their development.