Gigabyte X38-DS5 review & X38T-DQ6 revisit

November 6, 2007 | 13:43

Tags: #analysis #benchmark #crossfire #ddr2 #ds5 #ga #heatpipes #motherboard #overclocking #review #testing #x38

Companies: #ati #gigabyte #intel

Memory Performance:


Everest Memory Read Test

Everest Memory Read Test

  • Gigabyte GA-X38-DS5
  • Gigabyte GA-X38T-DQ6
  • Abit IP35 Pro
  • Asus Maximus Formula
  • 6786
  • 8309
  • 7534
  • 7320
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
MB/s

Everest Memory Write Test

Everest Memory Write Test

  • Gigabyte GA-X38-DS5
  • Gigabyte GA-X38T-DQ6
  • Abit IP35 Pro
  • Asus Maximus Formula
  • 6073
  • 6073
  • 6100
  • 6071
0
1000
2000
3000
4000
5000
6000
MB/s

Everest Memory Latency

Everest Memory Latency

  • Gigabyte GA-X38-DS5
  • Gigabyte GA-X38T-DQ6
  • Abit IP35 Pro
  • Asus Maximus Formula
  • 75.5
  • 61.1
  • 63.2
  • 66.8
0
10
20
30
40
50
60
70
80
nanoseconds (lower is better)

Unbuffered Memory Performance

Sisoft Sandra Unbuffered Memory Test

  • Gigabyte GA-X38-DS5
  • Gigabyte GA-X38T-DQ6
  • Abit IP35 Pro
  • Asus Maximus Formula
  • 6022.5
  • 6728.0
  • 6515.8
  • 5972.5
0
1000
2000
3000
4000
5000
6000
7000
MB/s

Memory Latency

Sisoft Sandra Random Memory Latency Test

  • Gigabyte GA-X38-DS5
  • Gigabyte GA-X38T-DQ6
  • Abit IP35 Pro
  • Asus Maximus Formula
  • 89
  • 79
  • 79
  • 84
0
10
20
30
40
50
60
70
80
90
nanoseconds

The motherboard product managers we've previously talked to about X38 DDR2 performance didn't hold back when they said it was "very poor" - both the Gigabyte DS5 and Asus Maximus Formula are outperformed by the older P35-based Abit IP35 Pro. The Asus seems to suffer less according to Everest, with a better read bandwidth and lower latency, but Sandra shows things on a more even footing. Synthetic write speeds are more dependent on front side bus performance rather than latency, and here 800MHz DDR2 or 1,333MHz DDR3 perform almost exactly the same.
Discuss this in the forums
YouTube logo
MSI MPG Velox 100R Chassis Review

October 14 2021 | 15:04