Go to the Penn State Home page

Go to the CLC Home page

Go to the ITS Home page
This site uses .Net links. Please use Text Only version for screen readers.  Text Only Printable Version    Secure Server Search CLC:   
   
  CLC Home
  News
  Labs
  Classrooms
  Assistive Technology
  Printing
  Disk Space
  Authentication
  University Services
  Contacts
  About Us
  Mission
  Staff Members
  Reports
  Projects
  Search

Benchmarks

While accurate benchmarks are difficult to obtain and results should be considered with suspicion, they are sometimes useful.

The Pocket PC Benchmark from Creative-software was used on the Pocket PC's, found via Benchmark HQ.  On the Axim, the processor setting was set to "maximum performance".  On all, all other applications were stopped and the device was on battery power and not docked.

Apparently (I'm not sure), for most numbers, smaller is better.  I'm surprised with the graphics benchmark; I had not noticed a difference between the Axim X30 and the others.  But the speed if the Axim X30 display is obvious when doing a few things like selecting "Programs" from the "Start" menu.  Displaying the screen does need CPU speed.

Model ID Siemens SX56 Dell Axim X30 Dell Axim X50 TOSHIBA e800
Free memory (Percent) 60% 40% 57% 41%
CPU Speed (mHz) 206 624 624 400
Loop count 3 3 3 3

CPU Benchmark
Integer arithmetic 9409 2486 2631 3205
Heap Management 11202 5401 5464 8161
Floatingpoint arithmetic 11159 4276 4344 6396

Graphics Benchmark
GDI graphic routines 1 2297 689 3305 2386
GDI graphic routines 2 2866 737 1271 1163
GDI graphic routines 3 1987 638 3656 1601

Flat File Benchmark
File: writing 9329 4438 4220 5841
File: writing 109.8 kb/s 230.7 kb/s 242.7 kb/s 175.3 kb/s
File: reading 7924 3276 4728 3807
File: reading 129227.7 kb/s 312576.3 kb/s 216582.1 kb/s 268978.2 kb/s

File Downloads

For these tests, the average of 3 lowest times, from 3 or more timings of saving a zip file from a web server to the device using IE, are listed.  The larger file was 5,000k bytes, or 39.0625 megabits.  For phone connections, the smaller file was 50k bytes or .390625 megabits. Actual data transferred would be more due to html and tcp protocol overhead.

Testing with 802.11 networks at the office were done after hours because our access points can be very busy during the day.  Timings with phone connections were more variable than other kinds of connections to the ranges are listed.

It is surprise that the downloads with the 100mb CF Ethernet card aren't much faster than 802.11b.  Is it the card or the device?

Device File Size Protocols Network Time (secs)
Dell Axim X30 5,000kB 802.11b w Movian VPN pennstate 30.9
Dell Axim X30 5,000kB 802.11b w/ 802.1x TKIP dot1xtest 17.4
Dell Axim X30 5,000kB 802.11b w/ WEP Home 16.3
Toshiba e800 5,000kB 802.11b w Movian VPN pennstate 39.3
Toshiba e800 5,000kB 802.11b w/ WEP Home 21.7
Toshiba e800 5,000kb 100baseT Ethernet Ethernet 18.3
Samsung i600 50kB 1X CDMA Verizon 11-53
Siemens SX56 50kB GPRS ATT 13-90

Data:

10/12: At home 10/12 (no VPN or 802.1x, but WEP encryption):
  Dell Axim X30: 16.3, 16.3, 16.4
  Toshiba e800: 21.7, 21.7, 21.6

10/13:
Samsung i600 phone: 15.6, 13.8, 15.2, 12.2, 12.1, 10.8, 15.6, 50.7, 52.6
Siemens SX56 phone: 64.4, 60.1, 14.8, 90.3, 42.7, 13.2, 58.7, 21.7, 21.9
Dell X30, dot1xtest: 18.4, 17.8, 18.0, 17.0, 17.4
Alternate between these two:
Toshiba e800, pennstate, 39.4, 40.0, 39.3, 39.3
Dell X30, pennstate, 32.0, 30.6, 30.7, 31.1
10/14
Samsung i600 phone: 43.4, 50.4, 52.9
Siemens SX56 phone: 17.2, 14.1, 14.8