Global MBT:
Login  |  Register          Free Newsletter Subscription
Email
Print
Reprint
Learn RSS

Network acceleration cures congestion for pharma supplier

By Staff -- Manufacturing Business Technology, 2/1/2008

Palo Alto, Calif.-based Affymax—a former division of Philadelphia-based pharmaceuticals giant GlaxoSmithKline—faced critical network congestion when it moved to a network-attached storage architecture.

The network, recalls Affymax Engineer Ben Maulis, had neither the speed nor the capacity to handle the volume of traffic the move entailed. As a result, full backups of the company's mission-critical information took four days.

The network-intensive analytical applications used by Affymax chemists and biologists didn't help—and neither did a “woefully inadequate” network infrastructure.

A plan to run 100Base-T cable to 250 desktops would help—as would installing Gigabit cable on the trunks and in the data center—but congestion at the server end still looked inevitable.

That didn't happen, thanks to Gigabit Accelerators from Alacritech, a company specializing in network-boosting acceleration technology.

"Within a server, the network and TCP/IP processing consumes more resources than people typically imagine."

—Doug Rainbolt,
VP, marketing & sales, Alacritech

“Server network and TCP/IP processing consumes more resources than people imagine,” says Doug Rainbolt, an Alacritech company VP. “Installing what can be thought of as turbocharged network interface cards designed to take responsibility for network and TCP/IP processing, and interfacing these directly to application memory, removes a lot of the network workload from the processor.”

Because the processor is no longer interrupted by network traffic, processor performance increases. And with network and TCP/IP processes handled by specialist hardware, network performance improves too.

“Compared to a standard network interface card, there's a 45-percent increase in throughput performance, but it can be as high as 300 percent, depending on the particular traffic being handled,” says Rainbolt.

“We successfully consolidated servers from 30 to 12, and significantly increased network throughput. Full backups now take only eight hours, a 75-percent reduction,” concludes Maulis.

Email
Print
Reprint
Learn RSS

Talkback

We would love your feedback!

Post a comment

» VIEW ALL TALKBACK THREADS

Sponsored Links



 
Advertisement

More Content

  • Webcasts
  • Podcasts
  • Blogs

Podcasts

Blogs

  • Mark Graban
    Get Lean and Prosper

    December 10, 2007
    Waste in the 787 Development Process
    Boeing Scrambles to Repair Problems With New Plane - WSJ.com Interesting article in today's WSJ...
    More
  • Mark Graban
    Get Lean and Prosper

    October 15, 2007
    Copying Toyota's Past Practices or Toyota's Thinking?
    With our Lean efforts, it's always good to ask, are we: Copying Toyota's past methods, ...
    More
  • » VIEW ALL BLOGS RSS
Advertisements





NEWSLETTERS
Plug in and get the latest MBT news, trends and industry updates delivered directly to your inbox!

Mid-Day Report (Twice Weekly)
Industrial Manufacturer Insider (Monthly)
MBT Europe (Twice Monthly)
White Space (Monthly)
Innovation Strategies (Monthly)
Intelligent Manufacturing (Monthly)
Lean Enterprise (Monthly)

About Us    |    Advertising Info    |   Site Map    |   Contact Us    |    FREE Subscription    |   Affiliate Links    |    RSS
©2008 Reed Business Information, a division of Reed Elsevier Inc. All rights reserved.
Use of this Web site is subject to its Terms of Use | Privacy Policy
Please visit these other Reed Business sites