Archives
 
 
 
  Special
 
 
 
  About Us
 
 
 

Newsletter
Free E-mail Newsletter from BYTE.com

 
    
           
Visit the home page Browse the four-year online archive Download platform-neutral CPU/FPU benchmarks Find information for advertisers, authors, vendors, subscribers Request free information on products written about or advertised in BYTE Submit a press release, or scan recent announcements Talk with BYTE's staff and readers about products and technologies

ArticlesSQL Server for NT on CISC and RISC


April 1994 / Special Report / SQL Server for NT on CISC and RISC

This test examines various hardware platforms for SQL Server for NT. The test machines were a SGI/ Mips Magnum (a 75- to 150-MHz R4400SC with 1 MB of cache memory and 32 MB of RAM), a DEC AXP 150 (a 150-MHz DECchip 21064 with 512 KB of cache memory and 32 MB of RAM), a Compaq Proliant (a dual 66-MHz Pentium with 256 KB of cache memory, and 32 MB of RAM), an Everex Step 486/50 (a 486DX2/50 with 256 KB of cache memory and 16 MB of RAM), and an IBM PS/2 Model 95 (a 486DX/33 with 256 KB of cache memory and 32 MB of RAM). The SGI/Mips Magnum is a favorite of mine. It's the fastest, smoothest NT workstation I've used regularly--a real thoroughbred. It's quickly becoming a collector's item, however, because SGI has passed the baton to NEC for the manufacturing of Mips-based NT workstations.

Waiting in the win gs was a Carrera Pantera II (a 200-MHz Alpha with 2 MB of cache memory) and an NEC Express RISCserver (a dual 75- to 150-MHz R4400), neither of which arrived in time for testing. On the horizon, DEC is working on Alpha-based multiprocessing systems.

The tests used are the query tests from NSTL. The transaction test results are not reported because of anomalies in the hard disk used for the transaction log.

About the Results

Best scores on individual queries were distributed rather evenly among the Compaq Deskpro 66/M, the Proliant, and the Magnum. The Proliant's multiprocessor technology, clearly capable of great results, can also falter. The Magnum's RISC engine similarly shows flashes of brilliance. But neither multiprocessing nor RISC clearly dominates these tests. RISC versus CISC? It's all just horsepower to NT, and if you can yoke the horses into teams, so much the better.


Windows NT SQL Server Query Performance Across Systems



                              IBM                COMPAQ                   DEC
                             PS/2     COMPAQ   PROLIANT       SGI/      ALPHA
                            MODEL    DESKPRO       (TWO       MIPS        AXP
                               95       66/M      CPUS)     MAGNUM        150


                              486        486    Pentium      R4400    DECchip
Processor                   DX/33     DX2/66         66         SC      21064
Processor clock rate (MHz)     33         66         66        150        150
Processor secondary cache  256 KB     256 KB     256 KB       1 MB     512 KB
RAM (MB)                       32         32         32         32         32


One user
Query 1                     10.43       7.82       4.24*      4.65      13.50**
Query 3                     15.46      12.87      35.37**     6.31*     14.07
Query 4                     28.22**    17.32       5.31*     13.68      16.41
Query 5                      5.30       2.89      13.22**     1.87*      4.38


Six users
Query 1                     21.16      16.38       8.89*     13.12      37.66**
Payment with query           0.48**     0.30       0.17*      0.22       0.34
Query 3                     39.86      31.51*    257.73**    62.29     174.47
ISBN order with query        2.91       1.49*      1.69       1.98       3.18**
Query 4                     40.16**    31.79       6.72*     18.61      25.37
Query 5                     10.39**     6.07       4.33       3.69*      9.08
ISBN order with query       12.93       6.80*      9.52       9.15      19.37**


12 users
Query 1                     44.22      34.97      20.08*     30.37      82.82**
Payment with query           0.86**     0.57       0.29*      0.38       0.60
Query 3                    112.11      76.15*    319.96**    87.20     273.75
ISBN order with query        6.36       3.12*      5.37       5.84       9.99**
Query 4                     74.47**    62.82      11.65*     34.25      46.53
Query 5                     17.33
      12.19      26.44**     6.16*     16.58
ISBN order with query       74.99**    21.16*     70.11      33.78      70.72


Numbers represent average response time in seconds.
* = best.
**= worst.


Up to the Special Report section contentsGo to previous article: How We Tested SQL Server on NetWare,OS/2, and NTGo to next article: LAN Operating-System TestingSearchSend a comment on this articleSubscribe to BYTE or BYTE on CD-ROM  
Flexible C++
Matthew Wilson
My approach to software engineering is far more pragmatic than it is theoretical--and no language better exemplifies this than C++.

more...

BYTE Digest

BYTE Digest editors every month analyze and evaluate the best articles from Information Week, EE Times, Dr. Dobb's Journal, Network Computing, Sys Admin, and dozens of other CMP publications—bringing you critical news and information about wireless communication, computer security, software development, embedded systems, and more!

Find out more

BYTE.com Store

BYTE CD-ROM
NOW, on one CD-ROM, you can instantly access more than 8 years of BYTE.
 
The Best of BYTE Volume 1: Programming Languages
The Best of BYTE
Volume 1: Programming Languages
In this issue of Best of BYTE, we bring together some of the leading programming language designers and implementors...

Copyright © 2005 CMP Media LLC, Privacy Policy, Your California Privacy rights, Terms of Service
Site comments: webmaster@byte.com
SDMG Web Sites: BYTE.com, C/C++ Users Journal, Dr. Dobb's Journal, MSDN Magazine, New Architect, SD Expo, SD Magazine, Sys Admin, The Perl Journal, UnixReview.com, Windows Developer Network