| 
  
	| OCZ Technology GameXStream 700W Power Supply Unit |  
    | 
  
    | Testing:
 
 To test the idle and load voltages of the OCZ GameXStream 700W, I installed the following equipment:
 
 
 
| Idle | Load |  
| AMD Opteron 148 processor @2200MHZ (stock voltage) DFI Lanparty UT NF4 SLI-DR Expert (default settings)
 1GB (2x512) Corsair TwinX1024-4000PRO DDR
 2x 80GB Seagate SATA 3Gbps hard drive in RAID 0
 1x 120GB Hitachi SATA hard drive
 Memorex DVD/CDRW
 10-n-1 Media Card reader
 eVGA 7900GT KO PCIe card
 Fans: 2x120mm
 | AMD Opteron 148 processor @2750MHZ (250x11) DFI Lanparty UT NF4 SLI-DR Expert (default settings)
 1GB (2x512) Corsair TwinX1024-4000PRO DDR
 2x 80GB Hitachi Deskstar SATA-II hard drive Raid 0
 2x 120GB Seagate 7200.9 SATA 3Gbps hard drive Raid 0
 1x 40GB Seagate 7200.8 SATA hard drives
 eVGA 7900GT KO PCIe card
 eVGA 7800GT PCIe
 Fans: 4x120mm, 6x80mm
 Cooler Master CoolDrive IV
 1x Hauppauge 350  PCI TV tuner
 2x USB thumb drives
 USB 2.0 Hub
 1x USB 2.5 hard drive
 Sony DRU-500A DVD-RW
 NEC DVD-RW
 |  The idle condition was established by powering the base system at default speeds and voltages with no applications running for a period of at least 30 minutes. The load condition was established by loading the system up with components and fans, while Folding@Home and SiSoft Sandra 2005's Burn-in Wizard ran for a period of at least 60 minutes. In addition to these stressful applications running, the loaded system was overclocked to add more demand to the power supply. The table below details the idle and load conditions:
 
 
  You can see both the idle and load voltages above, and that there was a slight drop in voltages between readings.  However, once the voltages dipped they did not fluctuate from these readings in the five (5) samplings I took.  The largest drop in voltages was on the 12V and 5V rails, and each drop was only 0.01; hardly worth mentioning.  These rails are the most stable rails I have ever used.
 
 The next chart compares the load voltage readings of four power supplies. All of the values are well within an acceptable range of +/- 5% of specification, and absolutely zero power issues were experienced during testing.
 
 
  With all the spare components I have laying around, I was unable to fully stress the 700W power supply.  Using various web power supply estimators, I was only able to get about 680W of components, not taking into account the overvolting and overclocking of the CPU.
 
 
 |  |  | 
 
 
 |