Internals, Testing Methodology and System Setup
InternalsAfter attacking the Colossus with a various assortment of power tools, the following resulted:
Inside we see a pair of cards plugged into a socketed daughter board.
These are SATA ports only in the physical sense. The electricals have been reworked to allow a pair of Indilinx SATA channels to pass through each single connector. In the above pic you can see a pair of data channels connected to each of two Silicon Image 5923 SATA RAID processors. The host side of each of these two controllers is in turn ganged together by a third 5923 processor on the flip side of the board.
Above we see a side of each of the two controller boards. Each board contains the equivalent of a pair of Vertex SSD’s. All together this amounts to a combined 256MB of cache and a whopping 64 4GB flash chips. The Colossus can scale as high as 1TB with current flash available as high as 16GB per chip.
(above paragraph edited to correct my poor attempt at doing math at 3 AM 🙂
(above paragraph edited to correct my poor attempt at doing math at 3 AM 🙂
Looking closer we see the popular Indilinx controllers paired with Toshiba flash memory. It’s nice to see Toshiba flash appearing in more devices, as this area was previously dominated by Samsung and Intel.
Specific for this review
All values will be obtained under XP X64 to keep the test data consistent with previous testing. Windows 7 was not required for these tests because the Colossus is not TRIM capable. The 4 Indilinx controllers *are* capable individually, but the the Silicon Image processor does not support TRIM forwarding so the controllers never get to see the TRIM commands.
Testing Methods
Our tests are a good mix of synthetic and real-world benchmarks. PCMark, IOMeter, HDTach, HDTune, Yapt and our custom File Copy test round out the selection to cover just about all bases. If you have any questions about our tests just drop into the Storage Forum and we’ll help you out!
Test System Setup
We have switched our testbed over to a more dedicated, forward thinking machine. We made some changes to help minimize test data scatter with higher bandwidth devices. Sound is flat out disabled, with no additional card installed. Video was *intentionally* shifted to a PCI unit to free up both PCIe-16 slots for testing tandem pairs of PCI Express cards (like ioDrives, DDRDrives, and high end RAID cards). Spot checks against the previous rig showed a negligible change in test output.
PC Perspective would like to thank ASUS, Corsair, and BFG for supplying some of the components of our test rig.
| Hard Drive Test System Setup | |
| CPU | Intel Core i7 920 @ 4 GHZ (HT disabled) |
| Motherboard | Asus P6T |
| Memory | Corsair Dominator 6GB DDR3-1600 |
| Hard Drive | G.Skill 32GB SLC SSD |
| Sound Card | N/A |
| Video Card | BFG Geforce 8400 GS 512MB PCI |
| Video Drivers | Geforce 181.22 |
| Power Supply | Corsair CMPSU-650TX |
| DirectX Version | DX9.0c |
| Operating System | Windows XP X64 SP2
|
- PCMark05
- Yapt
- IOMeter
- HDTach
- HDTune
- PCPer File Copy Test
