![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
So is SATA 150 a LOT faster then IDE66/100?
- Thread starter chay80
- Start date
Snugglebear
[H]F Junkie
- Joined
- Apr 25, 2001
- Messages
- 11,436
No. The higher interface speed lowers the overall latency between the buffer and the controller, but that's only a tiny fraction of the overall time consumed when the drive has to perform an operation. Outside of that your drive still operates at the same speed, which is still significantly slower than the interface. About the most useful trait of SATA right now, and the near future, is the cute little cable.
Icewindius
Gawd
- Joined
- Sep 30, 2003
- Messages
- 802
The only drive that truly takes advantage of the bigger headroom is Western Digitals Raptor series. The standard 7200RPM won't gain much of anything via SATA simpy due to the limitations of todays magnetic platter drives.
And you gain nothing from a SATA to ATA adapter, in fact, I hear it kills the lat timings
And you gain nothing from a SATA to ATA adapter, in fact, I hear it kills the lat timings
Snugglebear
[H]F Junkie
- Joined
- Apr 25, 2001
- Messages
- 11,436
If you run the signal through a bunch of adaptors, yes, you can kill the latency advantage. If you're dealing with a native set of components, the data will traverse the wire a few percent faster. That's still a pittance compared to how long it takes to swing the arm and platter around, though.
I
Ice Czar
Guest
there are 3 other advantages
1. Improved data integrity (better cable\standard)
Serial ATA and the 7 Deadly Sins of Parallel ATA
ATA not so Frequently Asked Questions
2. Hotswap support
3. Slight gain in efficiency
regarding frame information structure (FIS)
"If we take now a UATA-133 interface and device and transfer one singe block, the actual transfer rate can only be ~ 1/33 of 133 MB/s or 4 MB/s at the best. Averaves of 4 MB/sec will only be achieved if test patterns are used that employ repetitive single block transfers in back-to-back mode up to a certain transfer size. This data transfer rate is referred to in the industry as effective host transfer rate or TxH.
Using a SATA interface, the FIS overhead will be only ½ of what we see in a parallel interface and, therefore, the transfer rates can be as much as 1/17 of 150 MBs, at least in theory. With single block transfers, it also does not matter whether we are looking at single drives or RAID setups. "
in theory (there is very little realworld translation to performance) read the article (and the rest) @ Lost Circuits
As the Disk Spins (Stickied w\ sections, subsections and summaries at the top of the board)
1. Improved data integrity (better cable\standard)
Serial ATA and the 7 Deadly Sins of Parallel ATA
ATA not so Frequently Asked Questions
2. Hotswap support
3. Slight gain in efficiency
regarding frame information structure (FIS)
"If we take now a UATA-133 interface and device and transfer one singe block, the actual transfer rate can only be ~ 1/33 of 133 MB/s or 4 MB/s at the best. Averaves of 4 MB/sec will only be achieved if test patterns are used that employ repetitive single block transfers in back-to-back mode up to a certain transfer size. This data transfer rate is referred to in the industry as effective host transfer rate or TxH.
Using a SATA interface, the FIS overhead will be only ½ of what we see in a parallel interface and, therefore, the transfer rates can be as much as 1/17 of 150 MBs, at least in theory. With single block transfers, it also does not matter whether we are looking at single drives or RAID setups. "
in theory (there is very little realworld translation to performance) read the article (and the rest) @ Lost Circuits
As the Disk Spins (Stickied w\ sections, subsections and summaries at the top of the board)