Fermi Linux Server Vendor Qualification

1
Fermi Linux Server Vendor Qualification

• Project Closeout Report
• April 30, 2003

2
Executive Summary

• Technical evaluation of 1U dual Xeon units is
  complete
• 18 vendors identified as technically qualified
• Next step is price/performance bid. Requisition
  for 48 nodes in division approval now.
• Will weed down vendors to five.

3
Chronology

• We made contact with 45 vendors in all.
• 29 vendors attended Jan 28. info meeting
• 24 vendors submitted acceptable configuration on
  Feb. 4
• 21 vendors submitted acceptable benchmarks and
  were cleared to ship unit on Mar. 4all got it
  here by Mar 11.
• 18 vendors identified as technically qualified

4
Procedure

• Software check
• Most points evaluated from benchmark results
  vendors submitted by E-mail before units arrived.
• Hardware check
• All units opened, checked for quality
  construction and correct configuration.
  Pictures taken of all
• Electrical and thermal measurements
• One vendor disqualified for shoddy construction
  and bad service explanation of why it was that
  way.

5
Electrical

• Electric current measured with ammeter at
  startup, idle, and full CPU load.
• Current draw ranges 2.4GHz, 1.6-2.0A, 2.8 GHz,
  1.8-2.3A, 3.06GHz, 2.1-2.35A
• Likely that with purchase of 2.8 or 3.06GHz
  machines we can only have seven machines per
  circuit, not eight as in the past.
• Those with higher current draw also tend to have
  more fans and be better internally cooled.
• Bright sideThis current similar to 750MHz
  machines bought 3 years ago, 2.5x the performance
  for the same current.
• Table in the final report.

6
Thermal

• Measured ?T from front to back of unit for all.
• Used internal temperature probes on each unique
  type of case.
• All units in evaluation much cooler than the 1U
  units bought in FY2002.
• Due to better thermal characteristics of Intel
  chip and many more added internal fans and
  blowers.
• Still analyzing the data we took but confident
  that all units are acceptable.

7
Performance

• 3 CPU speeds measured, 2.4, 2.8, 3.06 GHZ,
• 1000 FermiCyclesPIII 1 GHz.
• Approximate performance, 1779,2041,2223 Fermi
  Cycles respectively.
• 400MHZ vs 533 MHz front side bus is 2.5 effect
  for farms software, much bigger for QCD.
• AMD MP2200 --1771 Fermi Cycles
• Performance is projected to faster clock speeds
  in anticipation that some vendors will bid faster
  chips.

8
Service

• All had at least one hardware call and one
  software call.
• One vendor disqualified because primary service
  provider listed wouldnt show up
• One vendor disqualified due to excessive downtime
  of machine.

9
Rack mount proposal

• Rackmount proposals evaluated for all
• Most still using conventional front-to-back
  cooling, standard rack.
• Several feature tool-free rails

10
Conclusions

• 18 technically qualified vendorsin alphabetical
  order
• Ace, Angstrom, APPRO, ASA, Aspen, Atipa,
  Concentric, Dell, HP, IBM, Koi, Penguin,
  Promicro, PSSC, Rackable, Racksaver, Richardson,
  Western Scientific
• Price/performance bid will weed them down to
  five.
• All technically qualified vendors will be
  remembered if some of the top five fail to
  perform, or if special capabilities, such as AMD,
  desktops, or blades, are needed.
• Requisition for 48 nodes is at division (32
  farms, 16 FNALU).
• 21 vendors is too many to bring in, will be more
  discriminating next time.

11
Future plans

• Next evaluation is an evaluation for blade
  servers, will start in late summer.

12
Thanks to

• Taskforce members Steven Timm, Margaret
  Greaney, Stan Naymola, Troy Dawson, Bruce
  Karrels, Lance Weems, Hans Wenzel, Bob Forster,
  Phil Lutz, Gerry Bellendir, Adam Walters, Mark
  Kaletka
• Other CD support staff (PREP, DCI) who helped get
  units in and out