Slashdot Mirror
Serial ATA Coming
John Doe writes "Heatseekerz.net Has a new article dedicated to Serial ATA @ Cebit 2002. This technology will be here sooner then you think!" The article is a little thin, but I haven't heard
a lot about what looks to be a very common standard in the not so distant future.
Find specs and other technical info here.
Until now hardrives have been limited to a master and a slave on a single controller. The Serial ATA standard allows you to connect more than two in a daisy chain similar to SCSI.
Hope that clears things up, it did for me.
[quote]
The reason for the late breakthrough of Serial ATA is that nobody wanted to, unnecessarily, spend time and money, developing a new interface.
[/quote]
Which seems like a good reason for not doing so.
I am in all likelihood missing something, but are we facing a bottleneck in ATA and ATAPI devices with their basic interfaces?... seems to be what we currently have has been able to improve and keep up.
Genuine question, not making rhetorical points.
It would seem to me, that unless Serial ATA offers clear and *necessary* benefits, it simply wont get support fro mthe chipset chaps.
I went looking for the benefits areticle on the cited page, but would appear they are struggling under their new load.
Q1: What is Serial ATA and Why is it being developed?
A1: Serial ATA is an evolutionary replacement for the Parallel ATA physical storage
interface. Serial ATA is scalable and will allow future enhancements to the computing
platform.
Q2: Previous efforts to transition to a serial bus were not successful. Why do you
believe that Serial ATA will be successful?
A2: Serial ATA is a drop-in solution in that it is compatible with today's software, which
will run on the new architecture without modifacation. It will provide for systems which
are easier to design, with cables that are simple to route and install, smaller cable
connectors, improve silicon design, and lower voltages which alleviate current design
requirements in Parallel ATA.
Q3: Will there still be a parallel ATA bus when Serial ATA comes out?
A3: Serial ATA's adoption by the industry will follow a phased transition path. There
will be a point where both Parallel and Serial ATA capabilities are available.
Q4: You stated that PCs implementing Serial ATA will be in the marketplace in
2002. Why does it take so long to implement?
A4: The goal of the working group is to ensure the Serial ATA transition happens as
smoothly and quickly as possible. The Serial ATA specification is expected to be
complete in the fall of 2000 with adoption to happen in the following 12 to 18 months.
Q5: What are the end user benefit of Serial ATA?
A5: End users will benefit by being able to easily upgrade their storage devices.
Configuration of Serial ATA devices will be much simpler, with many of today's
requirements on jumper and settings no longer needed.
Q6: What is the cost to implement Serial ATA in a system?
A6: The cost of Serial ATA technology will be on par with today's Parallel ATA
technology.
Q7: Who are the members of the Serial ATA Working Group? Can new companies
join?
A7: The Serial ATA promoters group includes APT Technologies Inc, Dell Computer
Corporation, International Business Machines, Intel Corporation, Maxtor Corporation,
Quantum Corporation, and Seagate Technology. Information on joining the working
group is available at www.serialata.org and new members are welcome.
Q8: Hard disk data rates don't seem to be pushing the limits of current ATA66
technology. Why is Serial ATA being planned now?
A8: Serial ATA is an evolutionary replacement for the Parallel ATA physical storage
interface and will allow future enhancements to the computing platform. Specifically, the
thinner Serial ATA cable addresses OEM's concerns regarding airflow around the
Parallel ATA cable, and enables design of smaller PC chassis, as well as silicon vendors
concerns regarding 5 volt tolerance support in future designs.
Q9: Will Serial ATA be compatible with today's PCs?
A9: Serial ATA electronics and connectors will differ from Parallel ATA, however the
technology is software compatible and OS transparent. It is anticipated that there will be
adapters to facilitate forward- and backward-compatibility of hard disks on PC systems.
Q10: What is the impact of Serial ATA on OEMs?
A10: Industry benefits of Serial ATA include systems which are easier to design with
cables that are simple to route and install, smaller cable connectors with improved silicon
design, lower voltage which alleviates current design requirements in Parallel ATA and
compatibility with today's software which will run on the new architecture without
modification.
Q11: Beyond hard disks, will Serial ATA be used on floppy drives, optical drives,
DVDs, and ZIP drives?
A11: Serial ATA supports all ATA and ATAPI devices, including CDs, DVDs, tapes
devices, high capacity removeable devices, zip drives, and CDRW's.
Q12: What is the impact of Serial ATA on IEEE1394 (aka Firewire) and on USB2
in terms of PC system function?
A12: Serial ATA is planned to be the primary storage interface inside the PC system,
and is not planned as an external interface to PC storage or peripherals. USB2 and
IEEE1394 connections on the PC can be used where required as peripheral interfaces.
Q13: When does Microsoft plan to support Serial ATA in its OS's?
A13: Serial ATA is software compatible with Parallel ATA and requires no changes to
Microsoft operating systems, or any other OS as well.
Q14: What are the licensing requirements and costs of Serial ATA to companies
that want to use the technology?
A14: When the Serial ATA specification is complete, it will be made available at no
charge. The working group expects to complete the specification later this year.
and way more informative... want specs? go here
"I do not fear computers. I fear lack of them." -Isaac Asimov
-replaces the long ribbon cable with a nice thin "serial" cable"
-replaces the seperate power connector and integrates it with the data cable
-standardizes the location of the data/power plugs
-allows for hotswap(partly because the location of the plugs are now standard)
For those who are still wondering about CPRM on the Serial ATA spec, these documents may be of some use.
USB2 and FW are no where near fast enough to be a next gen HD interface, nor where they designed as such. SATA is many times faster than FW/USB2 (500 MB/s vs 50 MB/s for FW) and supports key features for HDs (daisy chaining etc). USB2 and FW are nice general purpose external interfaces which also do an ok job with HDs, but they were never meant to repace IDE and never will.
(no pictures! :)
Serial ATA, A New Standard
Serial Advanced Technology A ttachment is an evolutionary high-speed serial link replacement for the parallel ATA attachment of internal storage devices. It connects hard disks, DVDs, CD-R/Ws, zip drives and all other ATA and ATAPI devices to the motherboard in desktop and mobile PCs, servers and network storage.
The new point-to-point device communicates trough a 4 layer interface:
4) Application Layer 3) Transport Layer 2) Link Layer 1) Physical Layer
If you want more information, you can download the Serial ATA 1.0 specifications at http://www.serialata.org
The reason for the late breakthrough of Serial ATA is that nobody wanted to, unnecessarily, spend time and money, developing a new interface. Because of the higher performance demands, several companies had to cooperate developing a much better, more intelligent ATA: Serial ATA will allow these future enhancements to the computing platform.
The Serial ATA working group that developed the Serial ATA specification was led by APT Technologies Inc, Dell Computer Corporation, International Business Machines (IBM), Intel Corporation, Maxtor Corporation, and Seagate Technology.
Seagate and Maxtor told us the new drives would be shipping in autumn, but the real breakthrough will take place when chipset manufacturers (Intel/Via) have integrated Serial ATA on their Southbridge.
Benefits
Let's take a look at some end user benefits of Serial ATA:
No software depency, it's 100% compatible with today's software and OS transparent.
Easier configuration of the storage devices (jumpers are no longer needed).
Supports lower cost device architectures.
Much better cabling and connectors: the thin and flexible cables result in better airflow trough the pc housing and enables design of smaller PC/motherboard chassis. Therefore, they are simple to route and install, and can be up to 1m long.
Last but not least, a higher bandwidth: the transfer rate exceeds all current ATA standards. Generation 1, 2 and 3 S-ATA supports respectively up to 150MB/s (1.2Gbits/sec), 300MB/s (2.4Gbits/sec), 600MB/s (4.8Gbits/sec). As Serial ATA works asynchronous, there are no isochronous requirements.
The Prototypes
Seagate's SATA prototype:
Maxtor's prototype:
When we take a closer look at the connectors, we see there's a slight difference: Maxtor used the 35B1 configuration, starting right: the Serial ATA connector, with both power and signal segments, legacy jumper and power connector. Seagate used the 35B4 configuration which has no legacy power connector.
The Serial ATA signal segment counts seven pins: three ground pins a transmitter signal pair and a receiver signal pair. The Serial ATA power segment counts 15 pins, containing three different voltages: 3.3V, 5V and 12V.
The prototypes momentary shown, need a PCI-to-SATA host controller or a SATA-to-Parallel ATA bridge chip.
Maxtor used the first one:
The Future
By the end of 2002, there should be SATA-sytems on the market. As you might have noticed, the parallel ATA is finally dead, though it could take up to four years to eliminata all parallel ATA devices. Nevertheless, I'm going to wait buying a new system, because most new technology suffers childhood disease.
-- null
Under Linux, each SCSI device still has it's own device driver, not a single device driver. Under Linux, there is a single device driver for disks just like there is under windows.
Linux isn't supperior in this, if anything, given the initrd method and the way in which ordering of driver loading is specific and the bugs in RH 7.2s installer, the MS method of signing drives and not changing their names depending upon the loading of SCSI drivers gives MS a bit of an edge in this storage model.
ATA is bus mastered, thats not the problem ... the problem is the lack of tagged queing causing lots of interrupts (or rather lack of support for it, its in the spec but noone is using it). SCSI drives also tend to have lower latency, but that is more an artificial differentiation by the manufacturers ... has little to do with the actual interface.
USB 2.0 is only 480M bits/sec -> best case Ultra 66 performance if you ignore USB 2.0 overhead
Firewire is at 400M bits/sec will be going to 800M bits/sec
Serial ATA is 1.2G bits/sec -> Ultra 150. There are little overhead on the protocol for data transfer.
Which one would the smart person pick ?
Why? Because Serial ATA is still ATA and SCSI is still SCSI. All the inherant benefits of SCSI are still limited to SCSI and all of the inherant drawbacks of ATA is still in ATA. Serial ATA doesn't really change the protocol, just the interconnects.
And Ultra320 is on the horizon already. And by the time Serial ATA reaches 600MB/sec I'd expect SCSI to have continued to grow.
Just because you can put two cows side by side doesn't make them an ox team.
If you think education is expensive, you should try ignorance -- Derek Bok, president of Harvard
However the Maxtor presentation talks about using SATA as a replacement for SCSI, or at a minimum breaking into NAS and low end servers.
Cable length is up to 3M from the Intel presentation.
Also for the number of devices, since it's now PTP connections, it's relatively agnostic as to the number of devices since it no longer fits into the old model of channels. I still haven't found specific references for the number of devices, but the Maxtor presentation has a picture of a SATA drop-in PCI card with 6 SATA connectors.
SCSI devices cant talk to eachother without utilizing CPU power either, one of the differences is indeed that data has to go through memory but the main difference is the lack of tagged queing support for ATA (its in the spec but noone supports it). You could even make an ATA controller with a local memory pool so the data could be copied locally instead of through system memory if you wanted to, luckily the CPU does not have to touch the data itself even with ATA.
Both SCSI and ATA have to be interrupted when the drives have to be told which blocks of data have to go where, the SCSI drive has no concept of a filesystem after all, but SCSI can que a whole lot of copy's while ATA has to be interrupted after each individual one has completed.
> Pardon my ignorance, but can somebody explain
> why serial ATA is faster than the current
> (parallel) ATA?
>
> On PC's, parallel ports are significantly
> faster than serial ports because they transmit
> 8 bits at a time instead of serial's one bit at
> a time. Wouldn't the same thing hold true for
> parallel vs serial ATA?
>
> Please explain.
In a perfect world, parallel would always be faster than serial. However, what happens is that due to outside factors (shape of the cable, EM interference at the time, etc) when you send those 8 bits down a parallel port, they don't all arrive at the destination at exactly the same time. The faster you send them, the more likely they are to not arrive when you send them since your tolerances get lower. This is referred to as signal skew.
Serial ATA borrows a technique from LVD SCSI devices which is low-voltage differential signaling. They send the pulses down 2 lines polarity reversed. By using 2 wires instead of 9 (8 data bits + a clock) or in UDMA land 20 wires (16 data lines, IOR, IOW, DMARQ, DREQ) the chances of them being significantly different than one another is less, because they're closer to following a consistent path through space. This allows them to toggle the lines MUCH faster in LVD applications than parallel applications, which gives us much higher data rates.
Another thing is that the IDE bus still uses TTL signal levels (5V/0V), meaning that it takes a HUGE amount of power to wiggle all those 40 conductors up and down to get some data across. The little chipsets on motherboards these days have trouble supplying enough internal power to do that, so LVD will help them make less complicated circuits in the chipset to talk serial ATA.
eric
More data, damnit!
> There are a number of issues that it seems that > SerialATA doesn't address that it should:
> 1) Power to the device is still separated from the data connection.
This is because the motors being currently used in hard drives pull a few amps at spinup time, and the wire guage used for signalling cannot possibly carry this much current. There is a thought that once the market is serial ATA native, the HD manufacturers will then standardize on a 5V, low current motor instead of the current 12V beasts, however, things like Microsoft's "on now" spec and other crap specify a minimum spinup time that force us to slam the motor to get up to speed.
Single connector vs two connectors has little bearing on whether you can hot swap. It is a function of how you isolate/protect your power circuitry from not having all the conductors touching at the same time. (e.g., in current ATA, what happens when of the +12,0,+5,-5, only the +12 and the -5 are connected because the pins are slightly out of tolerance?)
> 2) Because it is backwards compatible with
> regular ATA it appears it will have the same
> limitations on the number of devices you can
> connect, i.e. 2 per channel.
It is point to point. The notion of channels will disappear, and BIOSs in the future will simply allocate an 8-word I/O space address for the device, instead of todays "primary IDE" at 0x1F0 or whatever.
> 3) It is unusable for external devices
I don't believe that is correct, however, since it is point to point, a box of external drives (similar to a SCSI enclosure) would need a cable running to it for every drive in the enclosure.
There will be no daisy chaining or hub or star network of SATA devices.
More data, damnit!
There is a lot of misinformation being thrown around, so I thought I'd quote the spec:
2.1 Goals and objectives
Setial ATA is defined with the following goals and requirements listed in no particular order:
* Primarily inside-the-box storage connection (no outside the box)
* Completely SW transparent w/ ATA (easy transition)
* Low pin count for both host and devices (2 pairs)
* Favorable (low) voltages
* Supports lower cost device architectures
* Higher performance than equivalent ATA (data rate, queuing, overlap) w/ scalability to higher
* Much better cabling/connectors (thin, flexible)
* Includes efficient power delivery
* No software dependency. Relatively easy transition (price, IHV NRE and capital inventory risk, wide variety of devices at intro, etc.)
* Power management and power consunption suitable for mobile use
* Allows roadmap spanning ~10 years
* Cable length comparable to ATA (<1 m)
* Transfer rate exceeding best ATA (~150 MB/s) with scalability to higher rates
* Light protocol allowing overhead latencies to be minimized
* Asynchronous only (no isochronous requirements)
* No Peer-peer transfer support (to/from host only)
* Provides support for 1st party DMA access to host
* Cost competitive with equivalent parallel ATA solution at introduction (host + device + cable)
* Storage device centric (no cameras/scanners/printers)
* Easy installation/configuration (plug/play, no jumpers, no external terminators)
* Single host (no multi-initiators or host/host networking)
NO, it does not have $1/PC patent royalties per chip. Apple waived that years ago.
So there is not a price problem caused by Apple.
It's expensive because it's expensive. Because Intel invented USB, and Apple invented 1394, Intel has doggedly refused, even up to the present day, to support the standard on its own mobos. Intel FUD took care of the rest. So it has taken years to reduce the price of the chipsets -- but not because of the licensing fees. It's a matter of unit cost. Since Intel was actively hostile to the (superior) tech, it retarded the acceptance of 1394 and kept production costs artificially high. Chicken and egg...
But thanks to Apple, and common sense, people realized that Firewire was simply superior to USB in every way but price. Just compare an iPod to a standard USB MP3 player - transfer speed enormously faster, and Firewire also charges the iPod's batteries during the process.
So the prices came down despite determined opposition -- the market actually worked, sort of.
Firewired external CD-RW drives and hard drives work fine, and speedily. Putting the drive into the PC itself seems obvious.
The fly in this soup: 1394 developers seem determined to insert copy control into the cable/controller hardware. If there will be a choice between mediocre USB with no DRM, and 1394 with DRM, I'd throw the 1394/DRM equipment into the garbage, even if it were free!