Slashdot Mirror
eSATA Connectors
buffalocheese writes "Since the introduction of the Serial ATA 1.0a specification in 2002, many manufacturers have introduced PCI and CardBus cards with both internal and external SATA connections.
At first these internal and external connectors were completely identical, but later, external connectors started to appear which were still fully compatible with the internal sockets but featured added extra screening for external use.
With the introduction of the SATA II specification in mid 2004 a new external SATA connector was defined. These new external (eSATA) connectors are not compatible with the original internal SATA connection.
Currently there are add-on cards and drive housings available which feature both types of SATA connection for external use. Gradually the older types will disappear and all new SATA cards will feature the eSATA connector for external drive connections."
Frankly, I'm not a big fan of the current SATA connectors. They tend to fall off at the slightest provocation. I can't work in my case without having to check at the end that all of the SATA connectors are still in place (and at least one of them is usually loose or completely off). Because of this I've been reluctant to switch to SATA on external enclosures. If this new connector can prove itself resistant to falling off, it may in fact be a winner (I would even advocate unifying the connectors again under the new standard). I do like the fact that both the external and internal SATA connectors are currently the same, I just don't like the connector itself. It's rather nice to be able to take an old AT power supply (the kind where the power switch is hardwired to the supply) and plug in an off of the shelf SATA drive to the back of my case in a pinch. Plus, fewer connector types means fewer adapters I'll eventually have to own.
I read the internet for the articles.
I really don't see the advantage in having 2 types of connectors doing the same thing for internal and external use.
Except they want to sell me another cable - or did I miss anything?
While I'm not as certain when it comes to graphics cards, PCI-Express is/was sorely needed to replace PCI for general expansion cards. PCI had a limited amount of bandwidth available that was extremely easy to saturate (A single gigabit NIC would hit a bandwidth wall at something like 400mbit). The shared nature of the PCI bus was also an extremely frustrating limitation.
By contrast, PCI-Express 1x slightly increases the bandwidth from roughly 133MB/s to 150MB/s, but more importantly each device gets that, it's not shared anymore. And of course, 2x and higher slots provide more bandwidth.
But when it comes to graphics, AGP 8x was (at the time) providing more than enough bandwidth... as for the demands of modern monsterous graphics cards (such as the 8800 GTX), for all I know they might be able to saturate an AGP 8x bus.
As I understand it anyhow, the more tangible benefits from moving from AGP to PCI-Express were increase bi-directional bandwidth (AGP was great at Host->Card, but sucked at Card->Host), and increased ease with sticking multiple PCI-Express slots on the motherboard, making modern SLI possible.
One correction, one addition:
I'm fairly positive that it's 250 MB/sec per lane, not 150 for PCI-E.
In addition, not only is that per-device, it is per-device, per-direction (full duplex, 250MB/sec to the device and 250MB/sec back at the same time)
As to why PCI-E couldn't have been developed back when PCI or AGP were available (rather than incremental steps) - Moore's law. It simply wasn't possible to make silicon capable of handling PCI Express data rates (each lane uses serial communications at 2.5 gbits/sec, which was definately NOT possible with the silicon available back when PCI or AGP were initially developed.)
For those that wonder why PCI-E uses 2.5 Gbit/sec signaling but only transfers 250MB/sec of data, it is because all data is encoded using either 4B5B or 8B10B encoding (I can't remember which of the two), which maps every 4 data bits to 5 signal bits for 4B5B or 8-to-10 for 8B10B. This is done to ensure a minimum number of bit transitions in a given period of time, and also ensure that the signaling has no DC bias. (i.e. equal number of 0s and 1s no matter what the input data is).
retrorocket.o not found, launch anyway?
Oh, say does that Star-Spangled Banner entwine / The myrtle of Venus with Bacchus's vine?