Commentary
Penryn's Got The Power For Wall Street Computing
A fascinating nugget hidden amid Intel's announcement of its 45-nm Penryn processors is just who really needs these powerful chips. Sure, PC gamers want the hot Core 2 Extreme QX9650. And enterprises everywhere more or less buy into the "better performance per watt" sell of the server-side Xeon Penryns. But the folks for whom this stuff is like silicon heroin -- they can't live without it and they gotta have it now -- are the IT elite who run the data centers for the various stock exchanges.A fascinating nugget hidden amid Intel's announcement of its 45-nm Penryn processors is just who really needs these powerful chips. Sure, PC gamers want the hot Core 2 Extreme QX9650. And enterprises everywhere more or less buy into the "better performance per watt" sell of the server-side Xeon Penryns. But the folks for whom this stuff is like silicon heroin -- they can't live without it and they gotta have it now -- are the IT elite who run the data centers for the various stock exchanges.The reason is obvious: Today's stock exchanges handle hundreds of thousands of transactions per second, and their demand for ever-more processing power simply can't be satisfied quickly enough. Indeed, stock exchanges like NASDAQ in New York are to computing today what the national supercomputer centers were in the 1980s and 1990s.
Namely, in decades past, government money drove the procurement of and spurred the development of the world's fastest machines. As the twenty-first century began, though, Federal dollars dried up. These days, it's the ascendant electronic exchanges which take the best that Intel and AMD can dish out. They're also the proving ground for grid and advanced interconnect technologies.
More Global CIO Insights
White Papers
- Creating the Enterprise-Class Tablet Environment - by Yankee Group
- The BlackBerry PlayBook tablet's Good Bones - by BlackBerry
Reports
More >>Webcasts
- Maximize ROI with Database Consolidation onto Private Clouds
- Outsourcing Security: What Every Potential Cloud Security Customer Should Know
Take NASDAQ, which has seen a doubling in the number of transactions it handles in the past six months alone, according to Phil Marie, a senior vice president at the exchange. NASDAQ processes an astounding 124,000 transactions per second, of which 31,000 are stock trades.
Interestingly, NASDAQ is best able to do this by eschewing the old-style model of the centralized data center. Speaking at Intel's New York City Penryn launch event, Marie explained that NASDAQ has sold all its data centers and gone to a virtual data center model where they try to place all their facilities close to their customers. These satellite operations are then connected together via ultra-high-speed data links.
The objective: ensuring sub-millisecond execution rates.
So what we've got here is essentially a customer which demands real-time response in a hundred-thousand-plus (and growing) per second transaction-processing environment. Incredibly stressful, but also an incredibly important proving ground for every imaginable cutting-edge hardware and software technology out there.
Among that stuff, of course, are Intel's new 45-nm Xeons. Here's Intel's server processor roadmap:
![]() Roadmap for the server leg of Intel's 45-nm processors. (Click picture to enlarge.) |
Related Reading
| To upload an avatar photo, first complete your Disqus profile. | View the list of supported HTML tags you can use to style comments. | Please read our commenting policy. | |
|
|
T-Shirt Giveaway: Each week we're selecting one great comment from our readers. The author of the comment will receive an InformaitonWeek Community t-shirt. So get posting! |
Subscribe to RSSResource Links
This Week's Issue
Technology Whitepapers
- Mobile BI: Actionable Intelligence for the Agile Enterprise
- Creating the Enterprise-Class Tablet Environment - by Yankee Group
- How To Regain IT Control In An Increasingly Mobile World - by BlackBerry
- The BlackBerry PlayBook tablet's Good Bones - by BlackBerry
- New Visual and Wizard-Driven Paradigms for Exploring Data and Developing Analytic Workflows













