Shuttle SS21T

May 17, 2007 | 15:02

Tags: #am2 #barebone #benchmark #complete #evaluation #testing #xpc

Companies: #amd #asus #dell #shuttle

Memory Performance:


Unbuffered Memory Performance

Sisoft Sandra Unbuffered Memory Test (avg+int)

  • Shuttle SS21T (AMD FX62, onboard SiS Mirage 1 video)
  • Shuttle SS21T (AMD FX62, 7600 GT)
  • AOpen XC Cube EU965 (Intel X6800, onboard GMA950 video)
  • AOpen XC Cube EU965 (Intel X6800, 7600 GT)
  • Shuttle SD37P2 (Intel X6800, 7900 GTX)
  • Asus Crosshair (AMD FX62, 7900 GTX)
  • 5862
  • 6612
  • 5043
  • 5190
  • 5256
  • 8124
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
MB/s

Memory Latency

Sisoft Sandra Random Memory Latnecy Test

  • Shuttle SS21T (AMD FX62, onboard SiS Mirage 1 video)
  • Shuttle SS21T (AMD FX62, 7600GT)
  • AOpen XC Cube EU965 (Intel X6800, onboard GMA950 video)
  • AOpen XC Cube EU965 (Intel X6800, 7600GT)
  • Shuttle SD37P2 (Intel X6800, 7900GTX)
  • Asus Crosshair (AMD FX62, 7900 GTX)
  • 69
  • 67
  • 94
  • 90
  • 71
  • 78
0
25
50
75
100
nanoseconds

Whereas the memory performance is still in excess of Intel's Core 2 set ups, it is still far inferior to that of a typical ATX motherboard like the Asus Crosshair. Sure, the typical people who buy a Crosshair and SS21T are miles apart, but it shows the lack of optimisation the Shuttle suffers from despite having the exact same CPU and memory controller in both cases.

The SS21T is extremely limited by its lack of memory voltage adjustment, meaning that dropping the timings or overclocking the memory is almost impossible. We managed to reduce the timings down to 4-4-4-12 with the fastest, low profile memory we have: the Kingston HyperX PC9600, but CAS3 eluded us at just 1.84V. In comparison, even the AOpen XC Cube and Shuttle SD37P2 could manage 3-3-3-9 in its BIOS'.

The use of integrated video controller also cripples memory performance, sucking some 800MB/s (13 percent) of overall bandwidth for its own uses. Add this to the fact the output is so poor it hurts your eyes you're far better off investing in a cheap PCI-Express card.

Disk Performance:

The SATA read tests were conducted with a 74GB Western Digital Raptor using HD Tach 3.0.1.0's 8MB zone test. The USB 2.0 tests were done using an ATA133 160GB Maxtor DiamondMax 10 hard drive (model number: 6L160PO) using the same 8MB zone test enclosed in an Icy Box USB 2.0 hard drive caddy. We recorded the average read speeds in all cases.

SATA Hard Drive Performance

HDTach 3.0.1.0, 8MB Zone Test, Average Read

  • Shuttle SS21T (AMD FX62, onboard SiS Mirage 1 video)
  • Shuttle SS21T (AMD FX62, 7600GT)
  • AOpen XC Cube EU965 (Intel X6800, onboard GMA950 video)
  • AOpen XC Cube EU965 (Intel X6800, 7600GT)
  • 64.3
  • 64.3
  • 65.1
  • 65.1
0
10
20
30
40
50
60
70
MB/s

USB 2.0 Performance

HDTach 3.0.1.0, 8MB Zone Test, Average Read

  • Shuttle SS21T (AMD FX62, onboard SiS Mirage 1 video)
  • Shuttle SS21T (AMD FX62, 7600GT)
  • AOpen XC Cube EU965 (Intel X6800, onboard GMA950 video)
  • AOpen XC Cube EU965 (Intel X6800, 7600GT)
  • 28.1
  • 28.1
  • 33.2
  • 33.8
0
5
10
15
20
25
30
35
MB/s

SATA performance is very good, but USB 2.0 performance is a little low. It's not the lowest we've seen and you'd not expect a budget board to have the same optimisation treatment as enthusiast boards which battle for the best features and highest overclocks, requiring more engineering time.
Discuss this in the forums
YouTube logo
MSI MPG Velox 100R Chassis Review

October 14 2021 | 15:04