• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Tom's Hardware does Serial-ATA

Actually a very good article. I personally think I will wait until the newer southbridges have integrated S-ATA support, take it off the PCI bus, and give it some of the dedicated bandwith the high speed NB to SB connection provides...

🙂
 
I don't care about the speed, SATA won't push HD platter speeds any more than ATA 100 or 133 did. But those little cables are an anally neat case owner's dream.
 
Maybe it was an automatic spell checker, but Tom is saying Serial ATA Maxes out at 150 Mbps (megabits/sec).

At the same time, he says PCI is 133 Mbps.

Capitalize the "B," Tom. 🙂
 
The number of contacts on the power connector seems remarkably high at first glance. One of the reasons for this is that there is now the option of a third voltage as well as 5 V and 12 V; Serial ATA drives will now accept 3.3 V. 3.3 V is not normally available directly from the power supply. It will normally be supplied by the motherboard, which has to provide 3.3 V to a range of other components anyway.

Aren't pins 1 and 2 of an ATX power connector for the 3.3V wires? Or does this only apply to laptop drives where the power supply isn't standard ATX?

 
Maybe it was an automatic spell checker, but Tom is saying Serial ATA Maxes out at 150 Mbps (megabits/sec).

At the same time, he says PCI is 133 Mbps.

Capitalize the "B," Tom

Unfortunately this is a common mistake even in well established hardware sites...anyway Tom didn't do the article, one of his contributed staff named Patrick Schmid did it.

Edit: Having read the article, I didn't come across any instance where the author used Mbps, instead he used MBytes/s, which is actually correct and less ambiguous to casual readers.
 
Originally posted by: BDawg
Maybe it was an automatic spell checker, but Tom is saying Serial ATA Maxes out at 150 Mbps (megabits/sec).

At the same time, he says PCI is 133 Mbps.

Capitalize the "B," Tom. 🙂

Maybe they fixed it since you wrote that. It now says:

However, this elegant new standard does have its limitations. Serial ATA adapters use the PCI bus, which restricts the theoretical maximum data transfer rate of 150 MByte/s to the 133 MByte/s that the PCI bus allows.
 
Originally posted by: AmusedOne
Originally posted by: BDawg
Maybe it was an automatic spell checker, but Tom is saying Serial ATA Maxes out at 150 Mbps (megabits/sec).

At the same time, he says PCI is 133 Mbps.

Capitalize the "B," Tom. 🙂

Maybe they fixed it since you wrote that. It now says:

However, this elegant new standard does have its limitations. Serial ATA adapters use the PCI bus, which restricts the theoretical maximum data transfer rate of 150 MByte/s to the 133 MByte/s that the PCI bus allows.

Yep, they must have fixed it. I would think you proof your articles before you publish, not afterwards...or even worse, wait for a poster at another hardware site catch it. 😉

 
nice tech article, i would not hold any of the benchmarks to heart... but at least it seems that the performance loss from the adapter is minimal. some benches were higher...

also, that adapter from highpoint uses a mini-molex (floppy) power connection, this would mean we'd have to get adapters if all sata adapers use that. i recall seeing another adapter that was not in an enclosure from a few months back that used the 4pin molex. still, i dont like having 2 sets of cables go to the drive =\
 
Got two Qs about SATA so far, here goes.

How does SATA cables work, is it one drive per cable or 2+ like ATA and SCSI?

Will the format feature some ¤#¤&(U""%/% DRM system?
 
Keep in mind that none of these adapter cards, or the integrated controller on the IT7-MAX2 are true SATA controllers, but they use a chip to convert the SATA signals into Parallel ATA signals that are then controlled by plain jane ATA controllers. In the future (I assume by the time drives arrive) there will be true SATA controllers. I rather like the idea of a native controller as opposed to some translation chip.

Kramer
 
"Any chances that non-optical drives will go Serial-ATA as well? That way no more need for IDE cables. "

? Uhhh that doesn't make any sense. Everything to-date (including the article linked above) has been about non-optical drives (ie: harddrives). If you meant "Any chances that optical drives will go...." then the answer is yes absolutely, the whole point is to get rid of the old interface.

"The only thing that bothers me are the flimsy SATA connectors that don't latch on "

? Don't worry that will get fixed quickly, before the drives come out this fall.

"How does SATA cables work, is it one drive per cable or 2+ like ATA and SCSI?"

? Yes it's one drive per cable (point-to-point). However a controller (chip) can implement more then one interface (or port/cable) and it's up to manufacturers to decide which particular controller (chip) to implement. Check this thread and this thread for more information.

"Keep in mind that none of these adapter cards, or the integrated controller on the IT7-MAX2 are true SATA controllers, but they use a chip to convert the SATA signals into Parallel ATA signals that are then controlled by plain jane ATA controllers. In the future (I assume by the time drives arrive) there will be true SATA controllers. I rather like the idea of a native controller as opposed to some translation chip."

Completely true.

? As you can see in this picture has HighPoints usual ATA133 RAID controller (HPT372A) and two little Marvell chips however these aren't SerialATA controllers they're SerialATA to Parallel converter chips. So the Highpoint 1520 isn't actually a SerialATA solution it is currently a ParallelATA solution with SerialATA support***. (This would also apply to the motherboards you're talking about). This also explains why those adapters from HighPoint are so large ... when someone releases a real SATA product the adapters will be small dongles or gender-bender type devices.

? Check here for more info:
RocketRaid 1520 Review
(***See page 5 of the review for information on the PATA-SATA-PATA conversion latency etc...)

Thorin
 
"Serial ATA follows in the footsteps of other successful standards that are also based on the principle of serial data transfer. Ethernet, USB, FireWire ... "

Maybe Intel and AMD should convert their processors to serial instead of obsolete, outmoded, slow 32 bit parallel. Serial is so wonderful! Think of all the wires and transistors they would save if all the registers were 1 bit instead of 32. What a mistake the video card designers have made by going to 256 bit busses, when they could be using a 1 bit bus. What a terrible mistake RAID is, when you could have all the data coming from a single drive instead of several. What a terrible mistake it is to have dual or quad processors when you could have a single processor do everything.

We have 2.5GHz processors operating at 2.5 Hz X 32bits X 2 pipes = 160 gigabit rate. Think of how high it would be if Intel hadn't made the mistake of going parallel.

I guess we should be thankful the visionary HD experts are looking out for us.


"Thanks to serial transfer, serial ATA needs only two data channels - one for sending and one for receiving. ..."
A big mistake; they could have used one wire for both. I'm kidding. I guess they anticipate sending commands, or queing write data, while the HD is simultanously sending data.

"In short: connecting more than one device to a ribbon cable is a job we wouldn't wish upon our worst enemy. "
Oh yeah. It's horrible.

"...Serial ATA Working Group is now working on Serial ATA II and Serial ATA III, which will offer transfer rates of 300 and 600 MByte/s, respectively. "

So we can look forward to obsoleting two further generations of controllers, and maybe cables. How much are these new cables going to cost? Will they cost like the $25 USB cable you don't get with the printer? I think I know. Lets see, $25 for the cable, $60 for the controller card. $50 for the HD...


" Serial ATA drives will now accept 3.3 V. 3.3 V is not normally available directly from the power supply. It will normally be supplied by the motherboard, which has to provide 3.3 V to a range of other components anyway. "

Some mobos have a 3.3V out? Well, I guess we can get it through new controller card we will buy.

"Both, Intel's hub architecture and VIA's V-link, are basically nothing more than multiple, serial connections enabling bandwidths of 533 MByte/s with relatively few data lines."

Good point. All parallel connections transfer data in succesive groups, so I guess all parallel connections are multiply serial.


" Unfortunately, all storage devices still have to share the 133 MByte/s bandwidth of the PCI bus."
Aren't there other things grabbing chunks of the 133Mbytes/s already? USB? Sound? Video? LAN?

Realistically, since the old ATA HD spec was limited to 128G, things had to be revised. All the HD controllers are going to be obsoleted in the forseeable future anyway. So now is the time to change everything. Using the bandwidth of the wires more effectively by adjusting the transmission system makes sense. There are advantages to using fewer wires when that is possible, especially since everything else in the system limits the bandwidth to below what a single wire can provide.
 
sorry, I'm not as technically advanced as others in this conversation, so I've got a question. PCI bus supports 133mbyte/s, and serial ata supports 150mbyte/sec, so the PCI bus will be a limiting factor. But since this is a sustained rate, rather than burst, won't it still outperform current ATA133 drives?
 
Originally posted by: rugger29
sorry, I'm not as technically advanced as others in this conversation, so I've got a question. PCI bus supports 133mbyte/s, and serial ata supports 150mbyte/sec, so the PCI bus will be a limiting factor. But since this is a sustained rate, rather than burst, won't it still outperform current ATA133 drives?

Current ATA133 standards are capable of 133 MB sustained transfers, the limitation is that no ATA drive comes anywhere near 133 mbps sustained read - They can only achieve that on cache dumps - Which will never be above 8 MB, and 2 for most people since many do not have WD SE drives.

SATA will be the same way.

Viper GTS
 
"sorry, I'm not as technically advanced as others in this conversation, so I've got a question. PCI bus supports 133mbyte/s, and serial ata supports 150mbyte/sec, so the PCI bus will be a limiting factor. But since this is a sustained rate, rather than burst, won't it still outperform current ATA133 drives?"

Based on paper specs, yes the PCI bus will be the limiting factor for SerialATA controller cards. However since the best drive on the market currently averages 39MB/Sec (peak 49MB/Sec) it's not as if current ATA drives even challenge 66MB/sec (ATA66) let alone ATA133, SATA150 or the PCI bus' 133MB/Sec. However on the upside of SerialATA you can access more then one drive at a time so if you raided a few drives or transfered between two drives you could be sending multiple 39MB/Sec (peak 49MB/Sec) transfers at the same time.

Thorin
 
Back
Top