Fall 2003 IDF

Pages: 1 2 3 4 5 6 7 8 9 10 11

Personal Observations

While on the vendor’s floor, I managed to discover some interesting trends. One of the leading electronics manufacturers indicated that most of the demand they saw for IA64 based systems was from the smaller white box vendors who primarily sold standard HPC oriented solutions. However, I also spoke with two major OEMs that design their own proprietary IA64 chipsets and systems. Both indicated that they were unable to keep up with the demand for Itanium2 based systems, particularly among those running commercial server workloads. Another European OEM stated that they were beginning the conversion of their older mainframe clients to IA64 based systems, with their first trial already underway. While this may seem slow, it is important to remember that the customer base for commercial IA64 systems is incredibly conservative and unlikely to make any changes without a lot of feedback from their IHV and ISV’s. All together, this seems to be a good sign for the Itanium family; effectively, 2003 will be the first year that IA64 has really begun to sell outside of HPC, which is an important milestone for any new architecture.

Summary and Conclusion

This year’s IDF showed that Intel’s products are maturing as well as the corporate philosophy. The focus on technologies that do not immediately change performance, but improve the overall user experience, such as secure wireless communications, EFI and Trusted Computing, show that Intel and its partners are dedicated to using Moore’s Law to drive innovation into the end-user’s hands. Patrick Gelsinger observed that improving the user experience cannot occur without leaps forward in technology, and that it is important to make progress in tandem. Neglecting either the uses of technology or the technology itself will ultimately result in failure, and it is clear that Intel has diversified itself in order to provide both.

The immediate gains in performance from PCI Express, Serial ATA and DDRII are readily apparent, and the reliability and availability implications of virtualization will be one of the largest steps forward for mass market end users. However, the other technologies presented are just as important and may ultimately play a larger role in shaping the future; I would encourage anyone interested to read the now publicly available presentations, which can be found at Intel’s website.

References

[1] Hashemkhani, Mahnaz and Kitano, Jun. DDR2 Memory in 2004 Servers – Recipes for Successful Designs. Intel Developer Forum, Fall 2003.

[2] Findley, Geof et al. DDR2 Technology Readiness and Ramp Outlook. Intel Developer Forum, Fall 2003.

[3] Paddock, Bob. A Guide to Online Information about Low Voltage Differential Signaling (LVDS). Circuit Cellar Online. February, 2000. http://www.chipcenter.com/circuitcellar/february00/c0200r28.htm

[4] Interview with Mike Seibert, Strategic Marketing, Micron Technology. September 30, 2003.

[5] Grimsrud, Knut. Serial ATA 3Gbps and the Next Frontier. Intel Developer Forum, Fall 2003.

[6] Smith, J.E.. An Overview of Virtual Machine Architectures. 2001. http://www.ece.wisc.edu/~jes/902/papers/intro.pdf

[7] Privitt, Ken and Shim, Paul. Exploring Virtualizing and Partitioning on Servers. Intel Developer Forum, Fall 2003.

[8] Serial ATA II: Extensions to Serial ATA 1.0. SerialATA.org. October 10, 2003.

[9] Interview with Scott McLaughlin, Intel Corporation. November 4, 2003.


Pages: « Prev  1 2 3 4 5 6 7 8 9 10 11  

Be the first to discuss this article!