Serial ATA and Serial SCSI

aibrahim writes "In the recent Slashdot article about Serial ATA some people wanted to know where SCSI was going, and if Serial ATA could deal with some higher end workstation and low end server requirements. Apparently it has been decided that Serial ATA 2 (pdf doc) and Serial Attached SCSI are the answers."

Comments

Sounds like fun ...

[B3ryllium]

Now when can I get that 2.5 terabyte RAID1/0 array going?

Re:Sounds like fun ...

[HowlinMad]

screw that, I want raid 5!!!

Re:Sounds like fun ...

[slaker]

No. Probably you don't. RAID5 parity calculations pretty much kill write performance on every array I've ever seen. RAID5 gives a lot of bang for your buck, but if you want for-real disk performance AND redundancy, RAID10 is the way to go.

Copy/Paste a previously published article, anyone?

[jhoffoss]

Nearly anything is better than ribbon cables. aibrahim writes "In the recent /. article about Serial ATA some people wanted to know where SCSI was going, and if Serial ATA could deal with some higher end workstation and low end server requirements. Apparently it has been decided that Serial ATA 2 (pdf doc) and Serial Attached SCSI are the answers."

http://slashdot.org/article.pl?sid=02/07/04/153224 =nested=167

Difference

[SpatchMonkey]

The main difference is that Serial ATA will be more readily available first, and will therefore become more popular.

If you look at the Serial SCSI page in the FAQ, note that it is still under development, where motherboards supporting Serial ATA are out now.

Redundant story

[Danta]

Hey, this exact same story was just published in yesterday's slashback.

Wake up, Malda!

Re:Copy/Paste a previously published article, anyo

[Anonymous Coward]

Dammit. This one will work...
Article Here
(Oh yeah, this was posted yesterday... way to read your website, Taco.

I doubt it

[Jack Wagner]

I actually spent a bit of time on the SCSI board of standards back when they were commited towards making great strides in I/O throughput, and they have some real terrific advances that will be released in the next year when their patent's get approved. It's all hush hush and on the QT at the moment and my NDA runs through the end of 2003 so that's all I can say about it.

After briefly looking at the spec for Serial ATA I can see several limitations which will most certainly cause a bottelneck for any I/O simply due to the insane RPM's they require to actually hit their maximum VtR/mg (moving heads [sic] to magnetic resonence ratio). They need to have a look at perhaps doing some embedded hardware hacks which will allow them less resistance in the SCSI channel on the motherboard as opposed to quantum magnetic research. Just my opinion of course and not a reflection of the opinion of my company.

Best Regards,
-Jack

Re:I doubt it

[GigsVT]

They need to have a look at perhaps doing some embedded hardware hacks which will allow them less resistance in the SCSI channel on the motherboard as opposed to quantum magnetic research.

WTF?

A likely story, you sat on the "SCSI board", your consultancy domain name is a spammy squatter domain, you cite some BS ratio, you can't spell "resonance".

I may be making an ass of myself if you really do know what you are talking about, but I call bullshit.

Re:I doubt it

[Anonymous Coward]

For what it's worth: Jack was at one of the panels I attended January of last year giving a speak on extending circuit tolerance in solid-state SCSI drive circuitboards (the green board they wire the shiny bits of metal to at the bottom of hard drives these days). He wasn't a speaker, but I recall conversing with him and Rick Sterns from IBM's big iron division. Cool guy; could be making this shit up, but I doubt it.

Re:I doubt it

[alienmole]

You're right. Ignoring the quantum stuff, just the fact that he's complaining about I/O bottlenecks on a bus that's designed to support multiple devices is silly. The whole point is that a single device shouldn't be able to saturate the bus.

Re:I doubt it

[Anonymous Coward]

I took a look at that technical information - Serial SCSI certainly appears to have a very wide pipe.

firewire

[diesel66]

a good option to consider. i don't know of any mobo makers that offer it, but it's easy to add on. i think a speed bump to 800 Mbps is around the corner. include the fact that it's hot-swapable, 127 devices or something, etc...

Re:firewire

[charnov]

There aren't any Firewire native drives (that I know of...heh) and Serial-ATA should be cheaper. And, oh, be 1.5Gbs. Maybe now that Apple released all of the licensing surrounding the "standard", we will see it take off like it should have two years ago.

Re:firewire

[DansnBear]

I know of one company that has FireWire on their mobo. . . That would be Apple, the major company in the development of FireWire. heck, they even invented the name FireWire. If it wasn't for them, we would still be running around calling it IEEE 1394.

Re:firewire

[bofkentucky]

Lets look at it realistically:

• 800Megabits/s = 100MegaBytes/s, so transfer rate is the same as a standard IDE disk, after the speed bump.
• What's apple charging for a firewire license these days? Will impact pricing of controllers and drives?
• You mention hot-swapability, but most ATA drives aren't in hot swap capable enclosures.
• Is it just the interface or do other things need to change to allow hot swaping (More drive components = more expensive)
• Are there RAID 1+0 controllers for firewire drives?

Re:firewire

[diesel66]

ata100 doesn't use it's entire theoretical 100 MBs. also keep in mind that firewire's theoretical limits are 1600 Mbs (not MBs) (i s'pose it's still less than the 1.5 GBs of serial ata :) )

i think the license is $1 per controller/device.

hotswaping is built in. i haven't seen otherwise.

i was just thinking the cabling was cheaper than scsi and ata

Re:firewire

[Com2Kid]

The problem is that most Firewire HDs are just ATA HDs with a Firewire adapter shoved on them, thus they inherent all the limitations of ATA along with any limitations of Firewire. Doh.

Pain in the arse really, and why continue supporting two interfaces on one product if you don't have too?

Firewire is a good overall common denominator specification, but for something physically stable and performance based that also needs a low cost like consumer HDs, a dedicated standard is really best.

Re:firewire

[alienmole]

You're confusing different theoretical limits. The theoretical limit of current Firewire is 400Mbps. The idea is that the Firewire spec allows for future versions of Firewire to reach 1600Mbps, but current Firewire buses aren't even theoretically capable of that. In practice, 400Mbps Firewire doesn't get much above 315Mbps of actual data throughput - consistently slower even than ATA/66, which has been demonstrated in real-world tests.

OTOH, ATA/100 is theoretically capable of 100MB/s, i.e. 800Mbps, in its current incarnation today. You probably won't hit that in practice, not only because of bus limitations, but because 50MB/s is about the max any IDE drive can pump out, so you'd need two drives running continuously at their absolute peak speed in order to saturate the ATA/100 bus.

Re:firewire

[Anonymous Coward]

Actually, it was because of Apple that we were calling it IEEE-1394. They owned the name, and wouldn't let anyone else use it with royalty payments. It was only recently that they let up and have allowed the Firewire name to be used by IEEE-1394.

Re:firewire

[Anonymous Coward]

the new spec (1394b) has a theoretical limit of 3.2Gigabits (400MB). Whether anyone actually uses that remains to be seen.

Another post:
The royalty for current firewire cards is 25cents per card (U.S currency).

Lastly, the 1394 trade organization has legally adopted the 'FireWire' name with Apples' permission.

Re:firewire

[Anonymous Coward]

The most painful limitations of ATA are the master-slave stuff and the limit of devices per interface. Since FW-ATA adapters are 1 per drive, you don't have those problems.

Re:firewire

[DansnBear]

Apple owned the name "FireWire" just like Sony owned the name "iLink". IEEE-1394 was the established standard. Just because Apple and Sony wanted to impliment a piece of technology according a standard, and give it their own name dosn't mean that nobody else could do that too. Apple always gives seanly names to their technology. It's exactly like their useage of 802.11b technology. Find a standard, make it work, package it together in a pretty bundle, and give it cool name that everybody likes (AirPort) and people think your doing somthing special.

Sheesh!

[Mike Hicks]

...and just as I'm building my fibre channel array for my home computer :-p

Where are the drives?

[charnov]

Okay...I have a SATA equipped mobo on order which comes in in two weeks. What I want to know is: where are the hard drives? And no, I don't mean the drives that are standard ATA100 that have converters. I mean Seagates native SATA drives. They demoed supposed "production" pieces back in Fenruary.

Re:Where are the drives?

[bjschrock]

The Segate Baracudda ATA V, Serial ATA version is going to be released "this fall" according to a press release I read somewhere. It has a 8MB cache and comes in sizes up to 120GB.

Re:Where are the drives?

[alienmole]

Welcome to early adopterhood! Sounds like you need to wait for Seagate's drives in the fall, assuming they release on time.

You can bet the first SATA drives are going to be quite a bit more expensive than an equivalent capacity ATA100, too.

Time to start shopping for converted drives?

Re:Where are the drives?

[MrResistor]

You can bet the first SATA drives are going to be quite a bit more expensive than an equivalent capacity ATA100, too.

Not according to Seagate. I wish I could remember where I read that, but it was fairly recent.

Re:Where are the drives?

[alienmole]

I also read that initial drives were supposed to be comparably priced from the start. I just didn't believe it. :)

Re:Where are the drives?

[stu42j]

"the Serial ATA-based drives will likely carry just a $10 premium over their Ultra ATA/100 siblings, Eisman says."

From tech-junkie.com

as posted here.

It's funny how often slashdot readers ask questions that were just posted on slashdot. :)

Serial SCSI is neat.

[Sheetrock]

I've read a couple of brochures on this technology (we're looking at high-bandwidth high-availability file clusters for our hybrid AS/400-Solaris data warehousing) and it looks extremely promising.

Basically, they're extending parallel SCSI technology to address next generation I/O and direct attach storage requirements. It uses the (proven) interface from Serial ATA to avoid an unnecessarily proprietary interface and the costs that usually entails. The naming is unfortunate, because one usually thinks of parallel (side-by-side) as being faster than serial (one after the other) when the technology allows you to combine the two tactics much like in LANs. This is the technology that will enable a new generation of dense devices, such as small form factor hard drives, whereas Parallel SCSI can't because of cabling and voltage issues.

So depending on the pricing of the technology when it hits the shelves/junk mail catalogs, we're going to take a serious look at it. Does anybody have any prototype benchmarks?

Re:Serial SCSI is neat.

[Anonymous Coward]

The 1930s called, they want their lingo back :)

Re:Serial SCSI is neat.

[Anonymous Coward]

1. Serial SCSI development is just starting.2
2. OS400 drivers are only developed by IBM for storage at this time.
3. OS400 has never had an ATA driver of any sort in it.
4. The unit at IBM which designed AS400s custom SCSI controllers (like Gemstone) is under questionable status right now with the purchase of Mylex and Gemstone by LSI.
5. LSI and Adaptec are likely to be the only companies which produce early SAS parts, and they are over a year out for both companies.
6. There will *ALWAYS* be a price differential between high-end server storage and commodity storage.

Ultra-320 and most likely 640 will be the SCSI for the next 2-3 years in the mainstream.

Serial SCSI

[Anonymous Coward]

... It's called Firewire and within the next three years it'll be used to link up hard drives, networking and peripherals, according to Apple in 1996, unforunately in that time plain old ATA, Ethernet and USB have seen it off.

Serial ATA v. SAS

[bjschrock]

As the Serial ATA PDF points out:
Serial ATA is also expected to be a viable alternative for cost-sensitive entry-level and mid-range server and network storage applications. ... some of the advanced features of the SCSI protocol were not implemented.

So for some or most high-end storage applications, SAS and Fibre Channel will still beat out Serial ATA because you can do a lot more things with the SCSI protocol. Another big advantage of SCSI (in either SAS or Fibre Channel form) is the advantage of many targets to one initiator, or multiple initiators, with current parallel SCSI it's 15/1 and for Fibre Channel it's over 100 to 1 (I forgot the exact number). Multiple initiators aren't supported in Serial ATA II until phase 2, and Ultra 320 SCSI is already faster than the projected 300MBps of phase 2.

Re:Serial ATA v. SAS

[charnov]

Ultra320 SCSI is also INSANELY expensive...along with fibre channel.

Re:Serial ATA v. SAS

[Com2Kid]

I read someplace that ATA is just SCSI with all the goodies striped out to begin with, so this just sounds like a continuation of that pattern. ::shrugs:: Makes sense really, look at what the top performer is and how it works, and then take out all the very high end stuff that is not likely to be used with it in order to simplify it down and reduce costs, make it a bit more user friendly (at the cost of a bit of performance), and release it as a consumer spec.

Re:Serial ATA v. SAS

[iKitty]

Your performance assumptions are flawed. U320 SCSI allows 320MB/sec for the entire SHARED SCSI BUS (15 drives). S-ATA gives you 150MB/sec to EACH drive; the SATA controllers support a Direct Port Access (DPA) mode that allows DMA transactions to proceed simultaneously to all drives. With the new Seagate and WD drives having the larger 8MB cache in each drive we will see decent per drive performance and in large drive arrays (>8 drives) the performance should beat the pants off of SCSI.

Re:Serial ATA v. SAS

[FueledByRamen]

I beg to differ on this. Ultra320 SCSI may be insanely expensive; I've never used it, so I wouldn't know. However, Fibre Channel is not insanely expensive. If you look on eBay under a certain seller , 9.1gb 10,000RPM Seagate FC drives go for $9 each. A Fibre Channel interface card that sticks onto these drives and attaches them to the "loop," while normally costing $50 or more on eBay, I make for about $9 in parts each (including PCB). Cabling is simple - CAT5[e] works perfectly. The Fibre Channel HBA isn't too cheap - around $100 (this is the card you stick in your computer) - but I bought a pair of HP Copper HBAs for $20 each, shipped.

Re:Serial ATA v. SAS

[siliconwafer]

Very informative post, surely it should be modded higher than 2. Anyways, (being an idiot and not reading the article) does anyone know how many SATA devices can be attached to a single controller? Is it 2 like regular ATA?

Re:Serial ATA v. SAS

[Anonymous Coward]

so serial ata is faster than u320 scsi because the controller has more channels? do you realise that this makes no sense?

Re:Serial ATA v. SAS

[Anonymous Coward]

It will be faster (bandwith wise) in RAID systems because in practice the SCSI controllers will be too expensive and the cabling too bulky to only put 1 or 2 drives on a single port, with 2 drives per port it already has the same bandwith per drive as S-ATA ... anymore and it looses.

I believe..

[iONiUM]

I believe the general conception that SCSI is too expensive for the home user is going to make it hard for SCSI to pick up now. Although SCSI is much faster (and better for business in my opinion), I think IDE will continue to rule with it's slower perfomance and cheaper prices the home market. Quite a shame though, IDE seems to always be so slow when compared to the incredibly fast SCSI drives out... but then again, the size of IDE w/ current prices means you can get a huge hard drive for relatively cheap, which is almost impossible with SCSI.

Re:I believe..

[Courageous]

Well, you do know, don't you, that there is no reason why Seagate can't be selling those 15,000 RPM Cheetahs with an IDE interface if they wanted to, right? The SCSI bus issue doesn't significantly impact the drive speed in any way. It's a _BUS_.

C//

Re:I believe..

[abreauj]

the size of IDE w/ current prices means you can get a huge hard drive for relatively cheap, which is almost impossible with SCSI.

Actually, you can get an adapter that turns a cheap IDE drive into a scsi drive.
Chaeck out Acard's SCSI-IDE Bridge at http://www.acard.com/. I ordered one recently, and once I can squeeze a new hard drive into my budget, I'll be trying it out. Their FAQ says it handles IDE drives up to 128 GB.

Re:I believe..

[robertchin]

The general consensus is that SCSI is faster, but in reality, SCSI is only faster when there's more than one hard drive per bus. If you have only one drive per bus, IDE is faster than SCSI. So basically if chipsets supported four IDE buses so that you could separate each drive onto its own bus, SCSI would really not have any chance of being faster. This is what apple does on its XServe, give it a distinct price/performance advantage. If you really want speed, you go to FC-AL.

Internal firewire?

[Anonymous Coward]

While this is not important (hence my score:0 AC post) i would just like to ask a quick question related to this story topic i have been wondering about, as there tend to be many people on slashdot with more knowledge than i, so maybe someone will have an answer.

Does anyone know if there have been any steps anywhere in the Industry toward the eventual offering of internal hard drives that use FireWire? Would that not be cost efficient?

Re:Internal firewire?

[alienmole]

Current Firewire is half the bandwidth of ATA/100. Theoretically, Firewire can be saturated by a single fast disk, so for internal Firewire you'd want a separate Firewire channel on the motherboard for each disk - but Firewire was really intended to be a serial bus which supports multiple devices. That's why in its current form, Firewire is more appropriate for connecting devices like video recorders, or hooking up a single external drive for data portability, than for internal drives.

Re:Internal firewire?

[aibrahim]

IEEE 1394b supports bandwidth up to 3.2Gbps. TI has introduced a 1394b controller chipset.

Furtherless IEEE 1394b can run at 800Mbytes/sec over Cat 5 for up to 100m. Seems like you'll be able to just swap out your RJ-45 connectors for firewire ones and get to business.

As far as being saturated by a single fast disk...well do you have a single disk that can sustain 50Mbytes per second ? IEEE1394a can really transfer data very close to its theoretical limits in my experience. I've seen it shovel around 40+MBytes/sec so, I wouldn't write it off so fast. You need an IDE RAID 0 array to manage that.

Further, Apple is apparently considering rolling out 1394b as standard in the next round of desktops, and possibly the laptops too. (No link for the latter...)

Re:OSS...

[SirSlud]

I dont get it. Are these, like, 40 year old project manager trolls? Or, like, 19 year old scriptkiddie "Nike makes the best shoes, MS makes the best software" trolls?

I wanna know the demographic profile here that for some reason feels threatened by Open Source. Pure curiosity.

from the serial scsi FAQ

[Anonymous Coward]

# links -dump http://www.serialattachedscsi.com/faq/faq.html | perl -e 's/\ /\n/g' -p | grep leverag | wc -l
4
#

Four occurances of leverage or leveraging is too many in one FAQ. I don't trust them. I'm buying serial ATA.

Drop ATA

[Anonymous Coward]

Lets just drop ATA. It is an out of date spec. Lets just all buy SCSI, it will go down in price and will be as cheap as ATA soon enough.

just like the RAMBUS story

[lingqi]

remember back when RAMBUS said: we will provide an architecture with very narrow bus but extremely high speed to make up for it? (the *original* RAMBUS specs) -- beside the royalties and whatnot -- it actually (technology discussion only) had merits in that the PCB design was greatly simplified because of less crosstalk, easier routing, etc etc.

and then, people demanded more bandwidth... so now we have double / quad pumped RAMBUS channels -- in the end (today) it's back to 64-bit data-bus *anyhow*... except with an architecture that's not designed for parallel operation.

do anybody see some parallel (ha!) here?

i am guessing (or, predicting) that serial ATA / SCSI will go the same route. i really hope that it won't -- because if it did, our lives will all be kinda rough -- but it probabbly will.

sigh...

Re:just like the RAMBUS story

[jamesbulman]

Double / Quad pumped rambus channels refer to the frequency at which they are being driven, not the number of traces laid down. Rambus is as narrow as it has ever been.

Re:just like the RAMBUS story

[cheezedawg]

Ethernet has had no problems scaling to higher bandwith while maintaining its serial "bus". Serial ATA is a packetized interface that is more similar to ethernet than RAMBUS. They already have 600 MB/sec SATA on the roadmap.

Re:just like the RAMBUS story

[Anonymous Coward]

Actually, you are dead on target here.

One of the main limitations of sATA is that it is point to point. This means one cable per device. SAS will be similar, but out the door it will have the ability to use Hubs, where sATA won't until a later generation.

This means serial = more individual wires. So, it makes a lot more senese for IDE, where the standard configuration is only one or two devices today, and the cables are much smaller.

SERIAL (2D) versus VOLUME (3D)

[geekster_2000]

slow versus fast

http://colossalstorage.net/colossal.htm

Re:Can you imagine...

[Mode0x13]

Imagine a beowulf cluster of these posts.

Postercomment compression filter violated. Comment aborted == Add more CRAP!

Evil Drives

[Blahbbs]

So when will someone come out with a Serial ATA Network device?

Yup, a SATAN Device (tm).

...boo...hiss....

Right now

[Anonymous Coward]

Get 4 3ware (or Adaptec) 4 channel ATA raid controllers put em in a box (a really big one :) with 16 160 MB maxtor drives and you will have a 2.5 TB array for the measly sum of ~4000 dollars (performance will be quite good too, these controllers present a SCSI interface to the PC and use mostly hardware for the bookkeeping ... not like the Promise/Highpoint shit).

nope

[lingqi]

sorry for the mis-use of language -- what i am refering to is "dual-channel" RDRAM.

it was implemented on intel's i850 (? -- don't remember so well anymore) -- and required two modules to be installed simultaneously.

now that DDR / DDR II is catching up to RDRAM in terms of bandwidth, RAMBUS decided that all the "high performance" RDRAM modules will be "dual channel on a single chip" (which, btw, is 32 bit); now you will say -- this is still small -- but remember that originally RAMBUS can be used with only 8-bit bus width (somebody correct me if i got this wrong); and on the horizon quad-channel (64-bit) RAMBUS is looking at ya. guess how wide is the DDR / SDRAM bus? 64 bits too? ditto.

RDRAM is double-pumped (i do not believe it is a technical term, btw) -- data comes on both pos and neg edge of clock. there is no *real* quad-pumped memory; QDR-RAM is still only double-pumped except both I and O can operate simultaneously. only used in SRAMs anyway. (FYI)

Re:nope

[Fulcrum of Evil]

what i am refering to is "dual-channel" RDRAM.

The thing to remember there is that you don't have to synchronize the channels - they go as fast as they go. Multi-channel interleaved ram is a pretty easy way to speed up access, but it costs more (of course)

I heard

[Apreche]

that serial ATA, while being very fast and much better than what I've got now, will have DRM built in. Is this true? Should I not get serial ATA in my next system because of it? Anyone got any links pertaining to this issue?

Re:MOD PARENT UP

[Anonymous Coward]

you guys define pathetic

i think you are getting confused

[lingqi]

ETHERNET is a protocol; i think what you are really saying is that unshielded twisted pair was able to bump up in speed progressively.

i am not concered with the ATA protocol, in this case -- rather the amount of signals moved through the cables connecting the drive to your board;

even the venerable UTP can only get to 1Gbit and no more; ethernet lives on, 10Gbit ethernet standard is here, but guess what, fibre only.

same with ATA; you can only move so much signal (electrically) through wires. or, signals of so high a frequency; in this case, for a specific type of cable, there in a maximum amount of information that can travel through it. (unless you go out of your way to shield them, etc etc -- but a nicely shielded cable will cost you ~1500 dollars -- most high freq oscilloscope probes uses them, btw.) anyhow; serial ATA tries to bump up speed with a serial interface -- mainly to simplify MB design considerations -- less traces, narrower bus, etc; but since each strand in your cable will only go so far -- i am betting that eventually (without resorting to optical connections) even serial ATA, under the demand of higher throughput either by the market or by their (un)realistic roadmap -- start to double / quadruple the bus width. to me this is just silly -- because the benefit this offered is going away! MB designers will again have to fudge with wide busses and connections.

we might as well just keep on using parallel ATA but boost the signal freq incrementally, since it will get us to the same place in a few years anyway, without all these incremental MBs using different sized busses that's not compatible with different generation drives.

by the way -- PCI bus can only push 133MB anyhow -- anything beyond that is silly

Re:i think you are getting confused

[Anonymous Coward]

the venerable UTP can only get to 1Gbit and no more

But didn't they say the same thing when it reached 10Mb/s, claiming that's the limit and if you want any faster you'll have to go fiber?

Re:i think you are getting confused

[cheezedawg]

No, my point was they didn't have to add more signals to increase the speed of ethernet, and they probably won't have to with SATA either. Look at all of the hot technologies now: USB(2), firewire, 3GIO, SATA, and Serian SCSI. They are all serial interfaces.

we might as well just keep on using parallel ATA but boost the signal freq incrementally, since it will get us to the same place in a few years anyway

One of the biggest reasons for developing SATA was that parallel ATA is pretty much maxed out at 133 right now- we can't just "boost the signal freq" any more. Plus, parallel ATA is based on TTL signaling, and that requires the integrated circuits to tolerate +5V input signals. This is getting harder and harder to support with the modern manufacturing processes of the chip. And as you pointed out, the fewer signals also has the benefit of simplified design and reduced the cost of the chips.

by the way -- PCI bus can only push 133MB anyhow -- anything beyond that is silly

I hope you weren't serious about that.
#1. If PCI is becoming the bottleneck, then we will move past it. In fact, we already are. PCI's replacement (3GIO) is already in development (actually I think they changed the name to PCI Express- kind of dumb if you ask me).
#2. Integrated SATA implementations will not be on the PCI bus, so they will not be limited to the bandwidth of the bus. They will only be limited by the upstream bandwidth between the southbridge and northbridge chipsets (on current Intel desktop chipsets, this is 266 MB/sec with plans to increase to 533 MB/sec soon- AMD's Hypertransport is also more than adequate for SATA)

Re:i think you are getting confused

[lingqi]

USB(2), firewire, 3GIO, SATA, and Serian SCSI. They are all serial interfaces.

yes... look at a firewire cable; shielded twisted pair; the thing is, when you cram more and more signals through a cable, the cable itself gets quite pricy, so you are offloading design costs of the motherboard onto the quality control of cabelling, which i see no point in. high quality cables (shielded) cannot be bent too much because that will cause variations locally in the dielectric, screwing up your signal. and frankly, cables are much more likely to get bent / messed up than a trace in the PCB, so i rather see technology that's not very dependent on the cable quality / condition.

parallel ATA is pretty much maxed out at 133 right now

i am not denying this fact -- but at the mean time; i believe my argument still stands; what i do not like is the fact that suddenly, when moving to a new architecture, we decide that "we can do this with a narrow bus". i don't believe it. it is great and fine that you can reach the next step in your road map with only an 8-bit bus, but that does not mean you should do it. because i believe eventually the bus size will double and again. i hate to see that level of bs people will have to put up with. say i have a old MB when mainstream serial ATAs now have twice the bus width as when my MB was designed. how much are you willing to bet that it won't work then? USB and firewire will reach their maximum capacity (cable-wise) in the future, and when that happens, i bet you a dollar to a donut that the spec will start calling for wider busses on those too. but unlike specs that maintain their bus width, interoperability will be severly limited.

I hope you weren't serious about that.

i was about half serious -- and yes -- i know that SATA will be intergrated into the chipset directly, with it's own channels out. but wait a sec here... how does that simplify MB design again? you are saying i need a few more high freq traces going into the chipset, which is already crammed full of traces to the memory, AGP, southbridge, processor, etc? I would much more rather see a wider adoption of an evolutiona to the outdated PCI bus, and have things hang off those -- than have these new and fancy crap that gets crammed into the chipset. PCI, btw, *is* a bottleneck because any SATA adapter cards will hang off the PCI bus (say i want 8 drives for my system)... i wish the industry puts forth half as much momentum behind say, PCI64's adoption than SATA, etc.

Re:i think you are getting confused

[cheezedawg]

FWIW, the major OEM's have been clamoring for SATA because the SATA cables are cheaper than PATA ribbon cables.

And the 70+ companies in the SATA Working Group, the PCI-Sig (for 3GIO), and the USB and Firewire designers disagree with your assessment of the scalability of these serial interfaces.

And replacing the 26 (or whatever) signal pins that are currently integrated into the southbridge chipset for parallel ATA with the 4 signals for SATA certainly does simplify MB design.

Re:i think you are getting confused

[Spamboi]

No, my point was they didn't have to add more signals to increase the speed of ethernet

Ummmm, yes they did. There were two pairs used for 100BaseTX, a transmit pair and a receive pair. This left two pairs left over in a standard ethernet cable. For 1000BaseTX, the two extra pairs are used, and all the pairs are still clocked at the same 125MHz as fast ethernet. The tricks are that they use a more efficient encoding on the wire (giving two bits per symbol per pair), and, more importantly, that they simultaneously transmit and receive on the same pairs of wires, giving full duplex communication at 1 GBps clocked at 125MHz.

Anyway, my nit-picky point was that they did have to add signals to make gig-E work, in addition to using a set of crazy electrical tricks to make it work. I don't think it's fair to imply that Ethernet has near-infinite scalability over copper, or that any other serial protocol will scale infinitely either.

Andy

Re:MOD PARENT UP

[Anonymous Coward]

bite me

ATA/SCSI distinction

[XNormal]

With current silicon integration levels there is no real reason why SCSI should be more expensive than ATA. They could have just merged them and perhaps emulated braindead ATA on top of SCSI to keep compatibility or something if anyone really wants to.

I'm pretty sure the only reason they keep the difference is to be able to charge more from people building servers. It's purely a marketing and price positioning decision.

Re:ATA/SCSI distinction

[Anonymous Coward]

I agree. SCSI drives/technology has always been about creating a class system in the industry to squeeze more money out of peoples pockets. There is no engineering reason SCSI drives should cost more. These /. SCSI vs. ATA debates are a perfect example.

I think we all agree here that SCSI is a better engineered interface storage solution, but I for one am not willing to pay twice the price just so I can call myself some follow the Jone's superiority freak.

In many ways I'm glad that SATA may remove the class system and level the playing field for everyone.

SCSI...time to die.

Re:ATA/SCSI distinction

[Anonymous Coward]

There's no engineering reason that a low yield, high mtbf 36GB 10,000RPM drive with a 5 year warrenty would be any cheaper if it had an ATA interface -- Except that the target market wouldn't buy it and you'll find it in the discount bin.

Face it, the days of the same drive being produced with both ATA and SCSI interfaces are years and years ago.

The drive business is very competitive and the margins are ultra razor thin. Thinking that the nice server drives will become suddenly cheap if you changed the interface is retarded.

Aha! So I Serial Scsi is Good? Go IBM.. hmmm

[Anonymous Coward]

So when did IBM SSA (serial scsi architecture) which has done 160megabytes/s for the last few years suddenly become mainstream??

It didn't

Someone else has to 'reinvent' it. At least
the acronyn is almost the same. SAS vs SSA...

Hey why not! Reinventing old technology for the ignorant masses has always been Microsoft's Cash Cow.

back to the old days?

[PalmKiller]

So the serialized interfaces like on the old atari and commodore computers were ahead of their time, granted they were slow?

Re:back to the old days?

[Anonymous Coward]

>> So the serialized interfaces like on the old atari and commodore computers were ahead of their time, granted they were slow?

Not quite... ever changed the device ID on a Commodore 1541 floppy drive? Lets see... you power on the drives one at a time (with DeviceID #8 powered on last of course), and for each one you power on, run a complicated PEEK/POKE sequence to set the device ID prior to powering on the next drive in the sequence.

Alternately you can follow the instructions in the official drive manual to permanently set the ID by disassembling the monstrosity and *cut* circuit traces on the PCB and/or add a drop of solder to give you the desired device ID.

Ahh... memories.

Re:back to the old days?

[Anonymous Coward]

Better refresh your memories. It was a little messy and definitely hackish with the timed power-up sequence, but there were no peeks and pokes relative to the computer. You had to reprogram the *drive*, after all. A quick search will reveal several web pages with variants on the following:

OPEN15,8,15
PRINT#15,"M-W";CHR$(119)+CHR$(0)+CH R$(2)+CHR$(no+3 2)+CHR$(no+64)
CLOSE15

Re:back to the old days?

[Anonymous Coward]

Eeegads... getting technical over an ancient Commodore drive reprogramming sequence? How completely silly.

Why in this day and age would I even bother to perform "a quick search" and surf through "several web pages" to get the exact sequence unless I needed to actually do this? Heck it wasn't even worth the trouble to open the closet and quote the exact page from the actual 1541 manual I have stashed there, much less bother to waste the tiny amount of internet bandwidth the search would take. ;-)

Of course I'm bothering to reply as AC to an AC response to my AC reply so maybe I'm a silly one as well. :->>

S-ATA will dominate NAS

[Anonymous Coward]

The higher bandwith of Ultra-320 is not meant for a single drive, if you were to construct a system like that cost would go out of control. 150 MB/s per drive is plenty for any use, including high end servers.

If you have dedicated hardware talking to the drives things as command queing are totally meaningless (because it can just keep a queu itself and respond to interrupts as soon as they occur). So for big RAID/NAS with dedicated hardware the only thing SCSI has over S-ATA is the fact that the access time can be lower for single SCSI drives and sometimes higher reliability but that has fuck all to do with the protocol. The higher the number of drives the lower the average access time though, so with lots of drives this is not as big an issue as you might think, and you can just factor reliability of drives into the redundancy of your array ...

Areas will remain where SCSI rules supreme, but as long as the price differential exist they will get smaller and smaller. As clusters have shown, price always matters.

Re:Aha! So I Serial Scsi is Good? Go IBM.. hmmm

[chez69]

actually, SSA in IBM terms is serial storage architecture.

Re:MOD PARENT UP

[someonehasmyname]

I agree with this post . . .

So?

[gerardrj]

They still don't say that serial ATA will support more than two devices per channel. In fact they say it will be software compatible with ATA in its current form, suggesting it continues the master/slave relationship.

Today's drive media can only reach 40MB/s reading from the platters for short bursts, if their lucky. Normally they'll read/write about 20MB/s. What's the point of another boost in speed of ATA (to the suggested 150MB/s) when you will only ever be able to use 80MB/s of that. Oh, that's right... the ignorant users need bigger numbers on their cardboard boxes to show off to the neighbors.

Does anyone have any information on a HD soon to be released that will offer a quantum leap of read-from-meadia performance to something like 75MB/s? That's more than triple the current read-from-meadia speeds, and they seem to only ever increase the speeds by about 1-2MB/s each year.

SCSI makes sense having very high bus bandwidth, as you can connect quite a few devices and use the connect/disconnect to send simultaneous reads/writes to multiple devices. In that scheme, you can keep most of your drives operating at the same time. Of course Apple has shown that at least for a small RAID, multiple independent ATA channels are just as fast and lower cost than a single SCSI channel. I persoanally have a difficult time thinking that multi-ATA design would scale well to a 32 drive RAID, where a dual channel SCSI would shine.

Re:So?

[Anonymous Coward]

Waaahhh...crying about it doesn't help.

SCSI is no longer any faster than ATA.
SCSI is nearly twice the price of ATA.
SCSI disks are generally smaller than ATAs.

If you've got a PC, there's no reason to buy SCSI at all. If you do, you're just one of those ignorant users who need bigger prices on their sales receipts to show off to the neighbors.

Re:So?

[Anonymous Coward]

i can add to this and say that my tests have shown to me that with my setup of only 2 HD's in raid0, it doesn't matter whether I use multiple independent ATA devices or just stick them on the same ide cable.

Re:So?

[dallingham]

Serial ATA (SATA) is a point to point connection. Each SATA port can connect to one and only one drive. The whole concept of master/slave kind of goes out the window. Most SATA controllers will maintain software compatibility by emulating the ATA task file interface. The normal 4 drives can be supported without any special drivers, even though all the drives will be considered masters.

Yes, the current media rates are much slower that the interface rates. This is a much better situation than the other way around. Three years from now, when the drives are even faster, you'll be glad that the interface is that fast.

Internal buffer (cache) sizes in the drives are also getting larger. While this does not help sustained performance, you can really get a nice speed improvement if the data you want is in the cache on the drive.

Re:So?

[gerardrj]

But the buffers are insignificant. an 8MB buffer will be emptied or filled over the wire in .06 seconds at 133MB/s. So you can't get max throughput for anything longer than about .2 seconds.

Since the caches on the drives don't undstand filesystems or file structure, they can only contain things that have already been read from disk, or assume the next read will be a sequential block and pre-fetch that. More often than not, the cache on the drive does not contain the requested data. The disk cache only helps for small files that are re-read often like directories, and really the OS's disk cache will provide even better performance in these situations. The drive buffer does nothing to increase real-world data throughput on ATA disks, it's just there so the drive makers can claim a really high (wire speed) peak throughput number. Caches do make sense on SCSI drives where the drive can be ordered to read a set of blocks to buffer, disconnect, and later have the blocks read from buffer. During that drive's read phase (while disconnected) other drives can be commanded to read or write data to/from their buffers. This is why SCSI RAID systems outperform ATA RAID systems.

As for the increase in drive throughput from media: if future advances play out the way the industry has advanced in the past, it will be 15-20 years before a drive will be able to move 100MB/s sustained from rotational media. 10 years ago we where getting 10MB/s sustained, today we are getting 20, sometimes 30. Switching to some non-rotational media might see throughput increase dramatically, but all such devices I've seen connect to Firewire or USB[2] thus negating the need for more ATA bandwidth.

Serial ATA is a project in search of a problem, or perhaps more accurately marketing hype in search of consumer dollars.

Re:So?

[pangloss]

ok, i'll bite:

But the buffers are insignificant.

this is just theory, right? or do you have some benchmarks that you can link to?

i seem to recall that storagereview showed some marked performance improvements between the western digital xxxxBB and xxxxJB (aka "special edition") drives that supposedly only differ in the amount of cache (2MB vs 8MB).

Re:So?

[gerardrj]

I'm looking at a page on storagereview.com regarding the BB and JB models.
Summary:
Xfer rates: outer( BB: 49.3 JB:49.0) inner( BB:29.2 JB:29.2)
They don't mention the type of data they are writing, but it appears to be sequential reads/writes, probably in sector sizes. In any case the numbers are essentially identical. Looking at the 'desktop performance' page, there are wider, but still rather insignificant differences in performance. Certainly small enough that other factors could be causing some or all of the difference between drives.

Re:So?

[pangloss]

look at this one too, though. the numbers you were comparing are for two pretty different models--different capacities, platters, etc.

and here's a quote pulled from the above link:
"To differentiate their offerings from the competition, one of WD's largest OEMs recently requested an ATA drive with an 8-megabyte buffer. The manufacturer responded by retooling its current flagship, the WD1000BB, with an 8-meg cache."

which suggests that the cache bump was the only difference.

Re:So?

[gerardrj]

Okay... that result also shows almost no difference between the different models.

Serial Attached SCSI, redundant technology?

[Anonymous Coward]

Shoot me if I'm wrong, but isn't Small Computer Serial Interface (SCSI) ALREADY serial?
SCSI-1 was 8bit serial, SCSI-1 wide 16bit serial, SCSI-2 16bit serial SCSI-2wide 32bit serial, etc. but all serial. each bit channel independant of each other. Am I missing something?
I can understand it if they're just trying to get a small and manageable connector, but otherwise, considering the companies listed as members, they should know this already!

Serial SCSI?

[ghopper]

Isn't that called Firewire?

SCSI vs Firewire

Here is a better answer to both

[Anonymous Coward]

1394b supports up to 3.2gigabit and support for multiple devices and 100meter cable lenghts. It could be used to replace serial ata, ethernet,usb, and lots more.

here's an informative link.
http://standards.ieee.org/announcements/139 4bapp.h tml
It's real and it exists.

It will only support ONE device per channel

[Anonymous Coward]

No master, no slave ... just one lonesome cowboy.

It supports 150 MB/s because its trivial for them to support that speed, because there is only 1 device on the cable signalling becomes very robust. Also there's only 4 pins needed on an ASIC per port, so putting lots of ports on chips is no problem either ... all in all 1 device per port makes perfect sense, cables are cheap and no problem with drives contending for bandwith or needing to support disconnects in the protocol.

Re:It will only support ONE device per channel

[Anonymous Coward]

One port does not equal 4 pins, it equals about 10. Serial ports consume copious amounts of power compared to non-serial ports. Pinouts for early serial ATA devices do not show the reduced port counts.

The ATA relationship is still going to be kept in some ways. Given that sATA busses have severe I/O completion issues (basically resulting in just 1 I/O outstanding per device). SCSI can have hundreds of I/Os outstanding per bus and device, making it much more useful systems running multiple I/Os (servers).

But, bus speed is all about burst rates. Most drives can't sustain more than 50+MB/s at best. But for small I/Os, they can burst very fast. Latency on small I/O operations is what matters, and if you are doing quite a few of them, and getting cache hits, then a faster bus speed definitely is beneficial, as latency is reduced.

Re:It will only support ONE device per channel

[Anonymous Coward]

Ground pins are not an issue for pin count on the ASICs, neither are the pins for the new powersupply connector (which is entirely optional BTW, Im pretty sure most systems will rely on standard powersupply cables for drive power). Hard disks consume copious amounts of power period, two LVDS links are not going to make a dent in that either way.

ATA has a limited form of tagged queuing BTW, but for hardware RAID and NAS systems it does not need to be an issue ... it's easy to build hardware which can respond to service completion fast enough for it not to be an issue (so you keep the queu in the controller instead of the drive).

The Problem with SCSCI Today...

[evilviper]

My complaint about SCSI is the fact that it's split among so many different implimentations. That means vendors are going to choose to support one, and ignore the rest.

What we need is a single Serial SCSI standard (woo, try saing that a few times). Instead of Fibre Channel, Serial SCSI, Firewire, iSCSI, and whatever else they've come up with, we need one single interface for them all.

If it wasn't for dirvergent implimentations (25/68/80 Pins-Wide/Fast SCSI) people would most likely have SCSI in their systems today, instead of IDE/ATA. ATA retained backwards compatibility, while SCSI gave up compatibility just to get to market with a slight speed boost, as soon as they possibly could.

The other problem with SCSI, configuration and addressing difficulties, will not be an issue with Serial SCSI.

So I'd be happy if I could buy a Serial SCSI card with front-mountable Firewire ports (with more bus power than current Firewire), and perhaps with an option to buy an adapter if I wish to hook that card up to Fibre Channel devices.

Remember, this criticizm is comming from someone who *HATES* IDE/USB.

The SCSI groups REALLY need to unite on this stuff if they want to see any sort of advancement, rather than each ending up as a niche technology. Just look a the Alpha systems. I'm convinced they could have easilly replaced x86 systems. If SCSI groups keep going this way, they face the same ultimate fate.

Re:The Problem with SCSCI Today...

[Anonymous Coward]

"What we need is a single Serial SCSI standard..."? I can use an old scsi 2 drive on a scsi 3 system, I just use a good quality adapter. The whole I/O system takes a hit due to the older drive being slower, but I can still access it if I need to, so scsi is still backwards compatible.

"If it wasn't for dirvergent implimentations (25/68/80 Pins-Wide/Fast SCSI) people would most likely have SCSI..." niche solutions are just fine for the needs of those niches. There doesn't need to be any great one-size-fits-all solution for everyone. Have you looked at Microsoft lately, it isn't the greatest but works-well-enough solutions, oh and it is cheap too? Variety is a good thing, in hard drivers, operating systems, points of view, cultures, art, and so on.
Give the unaware, mee-too, marketing-like poster some nice new cement shoes, my big white toothy fish are really hungry!

Re:The Problem with SCSCI Today...

[evilviper]

I can use an old scsi 2 drive on a scsi 3 system

Well, you pointed out several drawbacks yourself, and besides, IDE/ATA drives do not need adapters, and don't take a performance hit for that compatibility.

niche solutions are just fine for the needs of those niches

Not true. The larger the group using any technology, the less money each needs to pay. Besides, those outside the niche are getting screwed as well (getting stuck with ATA/USB), not to mention being force to get new perhiperals every time a new interface overtakes an old one.

Have you looked at Microsoft lately, it isn't the greatest but works-well-enough solutions, oh and it is cheap too?

Microsoft is not the best around (as SCSI is), is not faster and better than the competition, and certainly is not cheap. Hell! If anything, Microsoft is the ATA/USB of software. It sucks, but it comes with every computer you buy.

Re:ATA/SCSI distinction: drivers

[Anonymous Coward]

Serial ATA is designed to be backwards compatible with older ATA drivers. Serial Attached SCSI is designed to be compatible with older SCSI drivers.

This may not be a concern for those of us running free operating systems, but it matters to many others. The ability to upgrade hardware even when your OS vendor doesn't support new standards is very nice. Throwing a new SCSI drive into my ancient SPARCstation 10 breathed new life into it. The new drive still supported SCSI 2. The same will be true for those running Windows who don't want to move to XP, if a new Serial ATA adapter still looks the same to the driver.

Re:OSS...

[Anonymous Coward]

MS does make the best software usually and certainly makes the best OS out there. Try not to embarrass yourself at each and ever chance you get, dullard.

Replacement for Hard Drives

[Anonymous Coward]

Let me know when we all get replacement for all hard drives.. too slow you can build the fastest system in the world but you will always be fucked by the hard drives. Let me know when i can load a 1 gig OS in .000001 secs. Otherwise serial speeds mean nothing... (unless you use it in a multi hard drive config)

For the love of GOD give use a new tech for storage.....