February 21, 2011
This week at ISSCC, Intel made its first detailed public disclosures about its upcoming “Poulson” next-generation Itanium CPU. While not in any sense complete, the details they did disclose paint a picture of a competent product that will continue to keep the heat on in the high-end UNIX systems market. Highlights include:
- Process — Poulson will be produced in a 32 nm process, skipping the intermediate 45 nm step that many observers expected to see as a step down from the current 65 nm Itanium process. This is a plus for Itanium consumers, since it allows for denser circuits and cheaper chips. With an industry record 3.1 billion transistors, Poulson needs all the help it can get keeping size and power down. The new process also promises major improvements in power efficiency.
- Cores and cache — Poulson will have 8 cores and 54 MB of on-chip cache, a huge amount, even for a cache-sensitive architecture like Itanium. Poulson will have a 12-issue pipeline instead of the current 6-issue pipeline, promising to extract more performance from existing code without any recompilation.
- Compatibility — Poulson is socket- and pin-compatible with the current Itanium 9300 CPU, which will mean that HP can move more quickly into production shipments when it's available.
Sadly, what Intel didn’t announce was the “real stuff” that people are interested in — availability and performance. While it's frustrating not to get this information now, its absence will not have any material effect on the acceptance of the product in the market. The vast majority of Poulson, like the current Itanium 9300, will be sold by HP in its HP-UX and NonStop server products, and the customers for these products have no urgent drivers that will make them abandon their platforms if Poulson doesn’t meet any particular date. The only sensitivity will be in new account wins, where the increased performance of Poulson will be a welcome competitive feature.
Will Poulson materially change the positioning of Itanium? Probably not. Intel is still (correctly) positioning Poulson as a processor for high-end niche environments, and the Xeon as the mainstream server chip. However, with each passing revision, the high-end Xeon CPUs will acquire more of the Itanium RAS features, and in combination with ever increasingly sophisticated versions of Linux (and Solaris, for Oracle customers), will continue to siphon current and future workloads from the Itanium. Xeon performance and price performance will also not lag, since by the time Poulson is ready to be manufactured in volume, the current 32 nm Westmere architecture will probably be available in its server variants.
So, in summary, Poulson is great news for HP-UX and NonStop customers, and not much of an issue for the rest of the industry. For current HP-UX and NonStop customers it means that they are virtually assured a 1.5X – 2X performance jump per socket within the next 12 months.