Originally Posted by Ktk
Absolutely true regarding the 400/250, but would that affect the other low benchmarks?
Of course, it will affect everything. First a Vertex 4 on a Marvell 9128:
Next same Vertex 4 on an Intel SATA III port:
All SandForce based SSDs do not do well in AS SSD, that is a given. But that is only in the write scores, they still have normal read speeds.
Any of the Marvell 91xx SATA chipsets are connected to one PCI-E 2.0 lane (hopefully PCI-E 2.0) whose speed is... 5Gb/s max. Is that the 6Gb/s of SATA III. No
, the end.
Note the write speed access times on the Kingston SSD. If that does not say "asynchronous NAND", the rest of the results do. (I know the M4's later firmware has even worse write latency, but we know that's the firmware.)
Asynchronous NAND is Single data rate
transfer at 66 MHz, while Synchronous NAND is Dual data rate
transfer at 100, 133 or 200MHz.
Or Async is 66MHz effective transfer rate vs Sync at 200, 266, or 400MHz effect transfer rate.
66 x 3 = 198. 172 x 3 = 516. I'm sure you all can figure out where these numbers came from and what they mean.