: 144GB Internal Storage"> Each HP BL20p blade can accommodate two Intel 1.4GHz Pentium III chips, two hot-pluggable SCSI drives, 4GB of error-correcting code synchronous dynamic RAM maximum memory and an integrated RAID controller with battery-backed write cache. The BL20p is also the first server blade to offer as much as 144GB of internal storage using two 10,000- or 15,000-rpm hot-pluggable SCSI drives. RLX systems can pack 324 blades in a rack but use processors that operate at half the speed of the BL20p and feature external storage only."One of the nice things about blades is that the cable count is significantly reduced, which improves reliability, and also the speed with which you can integrate these systems is faster," said Papadopoulos, who has worked with a beta version of HPs BL20p. Rather than one network and one power cable per 1U (1.75-inch) chassis, he said, the BL20p operates on one network and one power cable per eight blades, an 80 percent reduction. "Also, he said, "if something goes wrong with a particular machine, we can simply pull it out of the rack and replace it with a spare in a few minutes without having to take the whole system down." Larry Scott, at Dollar Rent A Car Systems Inc., said HPs beefier blades are better suited to meeting his business needs. "My interest is not just in total number of processors, but what is the compute horsepower in that unit of rack space," said Scott, manager of server support for the Tulsa, Okla., company, which uses HPs first blade product, the BL E-Class, introduced earlier this year. Dollar plans to deploy the new HP dual-CPU blades to help run its Web site, Dollar.com, as well as new Web servers, both customer-facing and internal, Scott said. Related stories:
Dual Chips to Power HP Blade Servers
Evaluation: Blade to Order
HP, RLX Sharpen Blade Servers
IBM Strengthens Ultradense Blade Server
Dell is New "Blade" Runner
Tech Analysis: Sharpening Server Blades
Blades offer several advantages over older server designs, said Philip Papadopoulos, program director for grid and cluster technology at the San Diego Supercomputer Center at the University of California.