Another blow to 1394?

Status
Not open for further replies.

EriMac

Seniorius Lurkius
42
Intel has really gone out of its way to try not to have anything to do with firewire. I guess its a pride thing related to USB. <P>Anyway, I still don't think the two necessarily compete with each other. I see USB as a natural successor to serial ports and as simply a well evolved I/O method for the personal computer. Firewire certainly has possibilities as a personal computer technology. But with the peer to peer networking being the real strength of firewire, as I see it is a lot more potential as we move into the era of the digital home, smart electronics, digital appliances etc. There are all sorts of futuristic potentials that firewire makes easy as you move into the world of the automated home where more and more electronics will be able to interface with each other to do astounding things. I'm not sure how long it will take for this to come to pass though. <P>Also 1394b is a lot closer at hand that USB 2.0. Personally I'm a little skeptical of USB 2.0 too. Firewire is trying to double its capacity from 400 to 800 Mb/s. USB 2.0 aims to get 20 to 30 times faster than current USB, 12Mb/s to 240-360. Thats pretty aggressive and if they can pull it off in the second half of next year I will be suprised. Many throughput technologies like SCSI, UDMA, AGP etc that I am aware of have evloved more at the pace of firewire which leads me to be a bit skeptical about USB 2.0 at least in the time frame that they are talking about. <P>And also, I have to say that firewire is just a damn cool name for a technology. View image: /infopop/emoticons/icon_biggrin.gif
 

total1087

Ars Scholae Palatinae
639
Heh heh...<P> <BLOCKQUOTE><font size="-1">quote:</font><HR>The USB peripheral bus standard was developed by Compaq, IBM, DEC, Intel, Microsoft, NEC, and Northern Telecom and the technology is available <B>without charge for all computer and device vendors</B><HR></BLOCKQUOTE><P>Poor resteves, getting his out-dated info all mixed up. And good thing Apple didn't have anything to do with it, since they'd probably charge $1 per port for USB.<P>Hey Apple guys, how do like using your USB stuff when Apple didn't have anything to do with the bus design? Especailly since M$ and intel had their hands into it, I guess you would be using "Wintel" technology on your Mac View image: /infopop/emoticons\icon_wink.gif <P>(sorry, cheap shot. Couldn't resist, tho View image: /infopop/emoticons\icon_wink.gif )
 

resteves

Ars Tribunus Militum
2,841
<BR>Gee total, not only an asshole, but condescending to boot...<P><BR>Where did you get your quote? From what I can find, those are the founding member of the USB implemetors forum, and they helped design the HCI guidelines, but they did not actually design the USB port. I am having trouble on the Intel site, so I could not get better information.<P>
 

Venture

Ars Legatus Legionis
21,830
"Hey Apple guys, how do like using your USB stuff when Apple didn't have anything to do with the bus design?"<P>Oh, there's plenty of original thinking going on at Apple over USB. The trouble is, it's original thinking about something that should be a standard.<P>If you design to Apple guidelines you have a product that works with maybe OS 8.6 but not OS 9, or with a G3 but not with an iMac.<P>And if you're a developer trying to develop Mac USB products (as a friend of mine is), Apple will tell you that 8.6 and 9, and G3s and iMacs all use USB in the same way, and why should they care if you're having trouble?<P>And when you spend three months getting the engineering team to come up with something (far removed from what Apple told you) that actually works across the board, at least you've got something that does more than a competing product (which appears to have been rushed out on the market and has trouble with anything other than 8.6 on an iMac).<P>What is most enlightening is that the market leader with around half of the market (and by far the widest range in this area) has not even bothered to implement USB in this section of its range. Are they waiting for Apple to finally come up with a reliable USB implementation? <P>
 

Venture

Ars Legatus Legionis
21,830
EriMac:<P>"Firewire is trying to double its capacity from 400 to 800 Mb/s. USB 2.0 aims to get 20 to 30 times faster than current USB, 12Mb/s to 240-360. Thats pretty aggressive and if they can pull it off in the second half of next year I will be suprised."<P>You might say the same for something like LocalTalk and EtherTalk; EtherTalk was a forty-fold speed increase in LocalTalk.<P>Methinks that the low speed of current USB was because it was intended as a replacement for serial ports. It wasn't until Apple started playing fast and loose with FireWire royalties that Intel became interested in USB2, and it's pretty clear that Intel doesn't want Apple to be able to hold the PC world to ransom.<P>It may be that the USB protocol is already viable for high speeds, and that the original 12Mb implementation was simply to provide a cheap way to mass adoption.
 

total1087

Ars Scholae Palatinae
639
Heh heh. Out of taking control of my work's B&W G3 and throwing SETI@HOME on it, and getting SETI on my Celery 450 on NT at home, I was bored and didn't like the way resteves posted. So, I "clubbed" him, if you will View image: /infopop/emoticons\icon_wink.gif<P>Anyway, I probably got him all upset and he might not answer back, such as he does whenever he's proven wrong. Oh well.
 

resteves

Ars Tribunus Militum
2,841
<P>I was bored and didn't like the way resteves posted. So, I "clubbed" him, if you will<BR>----------<BR>"clubbed" is an interesting choice of verbs. You quoted something, I asked you what you quoted because I could not find much to clearly support nor deny what you said, and you supplied it. I am clearly incorrect and, as always, am willing to admit it. I was under the belief that USB was developed by Intel and the others joined in to promote it, I did not realize that they had a hand in developing it. (Did they actually help develop it? Or did they just help design the general specs?)<BR>And what did you not like about my posting? That I calmly asked you where you got that quote? I never said you were wrong, just that I could not get good info regarding it. But I see you had to be a pompous ass and overcompensate again,<BR> <P> Anyway, I probably got him all upset and he might not answer back, such as he does whenever he's proven wrong. Oh well.<BR>----------<BR>I am rarely proven wrong, and when I am, I always come back and admit it. YOU seem to say I do it all the time, I challenge to to prove that assertion.<P><BR>
 

total1087

Ars Scholae Palatinae
639
Whoa, resteves. Calm down! The only time I didn't reply back to you was on Apple Sux, Linux Rulez. <P>Besides, this "suck it" and "clubbed" statements were all fun and games. That's why I put in the smiley faces. Heck, I'm a fun-lovin' guy, and was hoping you wouldn't get all offended at me by my posts. Live a little, k? View image: /infopop/emoticons\icon_wink.gif Besides, I really don't care who made USB. I just thought it was cute that M$ and intel made it, along with some of the major OEM peoples, and the irony that the Mac world advocates it so badly. View image: /infopop/emoticons/icon_smile.gif<P>Anyway, have a good Thanksgiving, k? And try not to let anyone piss you off like that. It's all just words, so as long as the facts come out. That's all I am worried about. No hard feelings, eh?<P>Oh, if you wanna club some baby seals, go here. And, I could care less if you think my bottom is pompous. Name-calling won't get you anywhere View image: /infopop/emoticons\icon_wink.gif
 
I thought that IEEE-1394 was engineered by Apple but the standards were developed by a consortium. Sony, Thompson SKG, Hewlett-Packard and others being in the consortium. This is the reason that the Chairman is from Sony, I thought. <P>Am I wrong? And I would like some info on this. Please point me in that direction. View image: /infopop/emoticons\icon_wink.gif
 

Venture

Ars Legatus Legionis
21,830
"I thought that IEEE-1394 was engineered by Apple but the standards were developed by a consortium. Sony, Thompson SKG, Hewlett-Packard and others being in the consortium. This is the reason that the Chairman is from Sony, I thought. <BR>Am I wrong? And I would like some info on this. Please point me in that direction"<P>AFAIK IEEE 1394 was from an original idea from Texas Instruments, the bulk of the work was done by Apple, and the protocol was submitted to the IEEE to make it a standard. Quite how the finished IEEE standard differs from what Apple submitted I don't know, but I suspect it's not by much.<P>There's also the IEEE 1394 trade association, which should not be confused with the FireWire patent pool. This last was formed after Apple announced that it would charge $1 a port for the use of FW. When this led to an outcry and Intel's rejection of 1394 in its chipsets, the other major DVD players (mainly the Japanese camcorder manufacturers) locked Apple in a closet and announced the patent pool and a royalty of 25 cents a port.<P>At one point it was said that FW2 would remove all the Apple code. Although FW2 was just announced I haven't seen anything that mentions this code.<P> <BR>
 

EriMac

Seniorius Lurkius
42
My understanding is that the 1394 patent pool which is known as 1394 Licensing Association (1394la.com I think?) handles licensing the technology to anyone who wants it. They also determine whether someone has a relevent patent and thus should be allowed to join their group to become a licensor.<P>The 1394 Trade Association handles the promotion and the specs and white papers and all that, basically everything but the money end of it. They can be found at 1394ta.org<P>As for history, I don't know whether the original idea came from TI or Apple or Apple&Sony jointly since I have heard all of those. But Apple did most of the early development work a while ago, in the mid to late eighties. Then it was presented to IEEE in the mid nineties. Some of the other companies like Sony and Compaq did work to refine various aspects for the IEEE standards (there are several variations: 1394, 1394a, 1394b etc) and that IP is what gave them the leverage to bitchslap Apple last year into forming the 1394LA to handle the $$$. Again, this is just my understanding of the facts, yours may differ depending on what you have read.<P>[This message has been edited by EriMac (edited November 28, 1999).]
 
How did this turn into another Apple vs PC rant? Oh yeah, its arstechnica forum...<BR>Anyway, the whole issue of USB vs 1394 is just another<BR>marketing FUD war started by Intel. USB 2 is a spec, nothing more. Until we see some hard evidence of chipsets and devices working in the OS space will we be able to gauge whether it lives up to it's claims.<BR>Some things to remember.<BR>USB will remain a host to peer spec, which basically means it is a "dumb" controller and requires a PC to achieve operability. <BR>IEEE 1394 is a peer to peer spec, which means that, for instance your digital camcorder can communicate with a hard drive without the need for a PC to drive the transfer.<BR>USB is dependent on CPU cycles to run, so like Intel technologies of the past the savings on cheap components is passed onto you, the consumer, with the ever-increasing need to buy newer more powerful cpu's. IDE anyone?<BR>1394 was designed from the ground up to be a simple, robust, high-speed interconect standard and is already specced to 3.2Gb/s with full backward and forward compatability and cable lengths of up to 100m. USB 2 is a complete redesign of the standard of Frankensteinian proportions and increased chipset complexity. In fact by the time it rolls out of the fabs, USB 2 will probably have a part cost very similar to 1394, though undoubtedly Intel will disguise this in order to maintain USB's "superior position".<BR>Bottom line. Everything USB 2 promises to do 1394 does better NOW, and still has capabilities not even on the USB agenda. <BR>If Intel had thrown their weight behind 1394, we could all be travelling towards a future with ONE interconnect standard and ridiculous simplicity, yet they choose to muddy the waters.<BR>Ironically, the Marketing Director for 1394 Trade Association is David Fair of, wait for it, Intel. <BR>Jeesus, talk about putting the fox in the henhouse.
 

Venture

Ars Legatus Legionis
21,830
"Anyway, the whole issue of USB vs 1394 is just another<BR>marketing FUD war started by Intel. USB 2 is a spec, nothing more. Until we see some hard evidence of chipsets and devices working in the OS space will we be able to gauge whether it lives up to it's claims."<P>So until we see some hard evidence (or rather, the lack of hard eveidence) of chipsets and devices we can't call it FUD.<P>"Some things to remember.<BR>USB will remain a host to peer spec, which basically means it is a "dumb" controller and requires a PC to achieve operability. <BR>IEEE 1394 is a peer to peer spec, which means that, for instance your digital camcorder can communicate with a hard drive without the need for a PC to drive the transfer."<P>Woo-hoo! Now if only I can find a HD that can read and write data at top FW speeds compared to top USB2 speeds.<P>"USB is dependent on CPU cycles to run, so like Intel technologies of the past the savings on cheap components is passed onto you, the consumer, with the ever-increasing need to buy newer more powerful cpu's. IDE anyone?"<P>IDE drives currently use around five percent or less of the CPU's resources.<P>"1394 was designed from the ground up to be a simple, robust, high-speed interconect standard and is already specced to 3.2Gb/s with full backward and forward compatability and cable lengths of up to 100m."<P>It was also designed to a big money-maker for Apple who got too greedy at the wrong time.<P>"USB 2 is a complete redesign of the standard of Frankensteinian proportions and increased chipset complexity."<P>No, it's faster. Can you tell me how SCSI went from 5MB/sec to 160MB/sec and somehow avoided these "Frankensteinian proportions"?<P>"In fact by the time it rolls out of the fabs, USB 2 will probably have a part cost very similar to 1394, though undoubtedly Intel will disguise this in order to maintain USB's "superior position"."<P>Since most of USB2 will be in the Intel chipsets, Intel will have to eat this cost.<P>"Bottom line. Everything USB 2 promises to do 1394 does better NOW, and still has capabilities not even on the USB agenda. <BR>If Intel had thrown their weight behind 1394, we could all be travelling towards a future with ONE interconnect standard"<P>We still are. It's called USB2. It might have been 1394 if Apple hadn't been greedy, and 1394 will still survive to connect camcorders.<P>But as an interconnect between computer peripherals, its future doesn't look too great.
 
"So until we see some hard evidence (or rather, the lack of hard eveidence) of chipsets and devices we can't call it FUD."<P>Dithering with their public support, then announcing a competing standard would amount to a fair amount of FUD on the part of Intel as is apparent by the ensuing news reports that spark discussions such as these.<P>Woo-hoo! Now if only I can find a HD that can read and write data at top FW speeds compared to top USB2 speeds.<P>Uh mate, USB2 doesn't exist. FW drive are available now, and by the time USB2 silicon theoretically rolls out in in the second half of next year, FW will be at 800Mb/s by the same assumption.<P>"IDE drives currently use around five percent or less of the CPU's resources."<P>So you mean the five IDE devices attached to my PC are sucking up to 25% of my CPU cycles. FUCK that.<P>"It was also designed to a big money-maker for Apple who got too greedy at the wrong time."<P>Big Yawn. Get over it. Its a stale argument, and an issue long since resolved. All the 1394 licencees are happy little campers paying the $0.25 royalty, working with an open ISO standards group. No, Apple don't have a thing more to do than any other board member with arbitrating 1394TO decisions.<P>"No, it's faster. Can you tell me how SCSI went from 5MB/sec to 160MB/sec and somehow avoided these "Frankensteinian proportions"?"<P>What are you saying? That over time SCSI has evolved into an elegant, simple and affordable solution? What planet are you from? 'Cause here on Planet Earth SCSI didn't fare to well, having a jumble of different connectors, termination complexities and compatability nightmares.<P>Since most of USB2 will be in the Intel chipsets, Intel will have to eat this cost.<P>Yah, Intel probably will, for the system manufacturers. For peripherals don't hold your breath. USB hubs will become increasingly complex and expensive due to the necessity of routing all the USB1 devices at differing speeds. This sort of thing was part of the agenda WHEN the 1394 spec was developed and as such, hub functionality is built into 1394 nodes.<P>"Bottom line. Everything USB 2 promises to do 1394 does better NOW, and still has capabilities not even on the USB agenda. <BR>If Intel had thrown their weight behind 1394, we could all be travelling towards a future with ONE interconnect standard"<P>"We still are. It's called USB2. It might have been 1394 if Apple hadn't been greedy, and 1394 will still survive to connect camcorders.<BR>But as an interconnect between computer peripherals, its future doesn't look too great."<P>Left my last comment there 'cause it looks like you didn't read it. More specifically, USB2 is incomplete in it's abilities and there is no roadmap to where (and if) it will go after this. Why wait for an unproven, at best mediocre standard when you can have a solid performer now. Some system manufacturers are already lining up behind the logic. Compaq, NEC, Sony, HP.<BR>Yeah not to mention some 10m or so 1394 camcorder owners looking for a socket.<BR>Looks like Intel misread the market again...
 

total1087

Ars Scholae Palatinae
639
<BLOCKQUOTE><font size="-1">quote:</font><HR>'Cause here on Planet Earth SCSI didn't fare to well, having a jumble of different connectors, termination complexities and compatability nightmares<HR></BLOCKQUOTE><P>I think the "What planet are YOU from" question should be asked of you. I have no clue where you got your SCSI standpoint, but SCSI is more of a powerhouse than it ever was. Especially since SCSI controller prices are going down, the simplicity of SCSI has always been around, and the DEPENDABILITY and RELIABILTY will ALWAYS out-weigh any other alternative (if you want to actually call them that, since I personally wouldn't say anything's an alternative to SCSI). <P>Besides, if you want to talk about transfer rates, SCSI is now at Ultra160, and has been at Ultra2 80. And since this is measured in MegaBYTES instead of MegaBITS, IEEE-1394 (<-- The REAL name of it) is MUCH slower. You do the math of 3.2Gbits and change it into Megabytes. And since by the time 1394 is up to 3.2Gbits, the SCSI standard will easily be past Ultra640 and beyond. And if you want to butt heads on current technology, check out Fiber Optic SCSI hard drives and controllers. 200MBytes, 126 devices per connection, up to 10km of fiber optics between each device, etc. <P>But, if one would check out your post, you sound like some 1394 Advocate (read: Mac person-type) that's trying to pick a fight when there's nothing to fight about. Heck, we've already acknowledged the advantages and dis-advantages of both said controllers. We know USB sucks, USB-2 is curently vapourware, 1394 sucks (still in its infancy), and if other companies don't adopt it soon and push out its 1394 products, it won't fair well, either. But, it doesn't really matter. <P>In any case, the new Mac's are using M$ and intel technology anyway (read: USB). Actually, there is nothing Apple about any of the components inside their new systems, except for the colorful plastic shell that's holding third party parts. Oh well.<P>Hmmm... I wonder when Apple will finally have their own FAB plant(s) like the rest of the (computer) industy...
 
PimpiumIII:<BR> <BLOCKQUOTE><font size="-1">quote:</font><HR>So you mean the five IDE devices attached to my PC are sucking up to 25% of my CPU cycles. FUCK that.<HR></BLOCKQUOTE><P>Umm, I gotta start out with a "duh" here, because I don't think you understand the background (your post had the tone of a defensive Mac'er, at least that's how it sounded to me). What the original quote was saying is that IDE drives take up 5% CPU time for doing disk activity. If the drive isn't being accessed, it's not sucking up 5% CPU time. DMA controllers consume very little CPU time, and all current PCs have DMA IDE controllers built into them (and have for quite some time).<P>And NO, having 5 IDE drives doesn't mean that 5% is accumulated for EACH drive. Besides that, how do YOU have five IDE devices in your computer? Have a DMA66 controller in there along with the standard DMA33? Cuz normally you're only gonna have 2 IDE controllers in a computer, which each control 2 devices (and 2+2=4).
 

John

Ars Praefectus
3,788
Subscriptor++
<BLOCKQUOTE><I>[IEEE-1394] was also designed to a big money-maker for Apple</I></BLOCKQUOTE><P>Hey, they're not a non-profit organization! At least they gave up the specs to IEEE, which is more than Intel has done with USB. If the situation was reversed, people would be screaming to high heaven that "Apple is trying to foist its proprietary, Apple-controlled, non-existent USB2 hardware standard on the masses, displacing Intel's open IEEE-1394 standard."<P>Don't kid yourself, USB2 is a play by Intel to get another part of the PC under its control. In a contest between even an open IEEE standard and something totally controlled by Intel, guess which one Intel is backing? It has nothing to do with technology. Furthermore, IEEE-1394 has already been chosen by the consumer electronics industry to be the interconnect standard for digital video and HDTV. So if USB2 does win the war for the high-speed serial interface on the PC, we will have squandered yet another chance to get the CE and PC worlds onto the same interface...which will be even more of a shame with the advent of digital TV.
 

EriMac

Seniorius Lurkius
42
Perhaps you could clarify a <BLOCKQUOTE><font size="-1">quote:</font><HR>In any case, the new Mac's are using M$ and intel technology anyway (read: USB)<HR></BLOCKQUOTE><P>And what is the relevence of that?<P> <BLOCKQUOTE><font size="-1">quote:</font><HR>Actually, there is nothing Apple about any of the components inside their new systems, except for the colorful plastic shell that's holding third party parts.<HR></BLOCKQUOTE><P>Do you mean designed by Apple or manufactured by Apple? <BR>
 

resteves

Ars Tribunus Militum
2,841
<P>In any case, the new Mac's are using M$ and intel technology anyway (read: USB). Actually, there is nothing Apple about any of the<BR> components inside their new systems, except for the colorful plastic shell that's holding third party parts. Oh well.<P> Hmmm... I wonder when Apple will finally have their own FAB plant(s) like the rest of the (computer) industy...<P>----------<BR>And exactly what does gateway/Dell/Compaq fabricate? Heck, what do they design? Apple does make some of its own stuff, but it designs a lot of it. The simple stuff (cd roms, HD, etc) are off the shelf, but the motherboard, ASICs, etc are by apple design. (I don't know if they fab them or not.)<BR>
 

total1087

Ars Scholae Palatinae
639
John:<BR> <BLOCKQUOTE><font size="-1">quote:</font><HR>At least they gave up the specs to IEEE, which is more than Intel has done with USB<HR></BLOCKQUOTE><BR>I think you may be a little mis-informed. Apple themselves state that intel doesn't control USB, but the stated companies do, AND the technology is free for all computer and device vendors.<P>EriMac: The reason I brought up USB as M$ and intel technology (again) was in accordance to PimpiomIII's "Looks like Intel misread the market again..." statement, among other things he said concerning intel (read: bashing intel). I wanted to clarify that if it wasn't for the heavy hitters such as intel and M$, USB wouldn't be around for his (future?) enjoyment. <P>Also, my statement <I>"Actually, there is nothing Apple about any of the components inside their new systems, except for the colorful plastic shell that's holding third party parts"</I> was in accordance to manufactured and designed by Apple. But, resteves was kind enough to point out that Apple does manufacture their own motherboards (which apparantly slipped my mind; thanks for correcting me, resteves!).<P>resteves: Are you aware of FAB plants and what they manufacture? I double'checked my work's G3 and the 8500, and the only ASIC's in these machines are made by TI and Sharp. There is absolutely NO Apple silicon on the G3 motherboard and the 8500 motherboard. That's what I mean about Apple not having any FAB plants, since they don't manufacture any silicon themselves, but 3rd party companies do (IBM and Moto).<P>Oh, and Dell/Compaq/Gateway DO have their own FAB plants (but, not the size of intel and AMD and the like). They used to manufacture their own CPU chipsets and video chipsets in the day, and I do know Compaq and Dell still use them for their Enterprise Servers (ASIC's and stuff like that). <P>[Edit: used BOLD instead of italics]<P>[This message has been edited by total1087 (edited November 29, 1999).]
 

John

Ars Praefectus
3,788
Subscriptor++
<BLOCKQUOTE><I>I think you may be a little mis-informed. Apple themselves state that intel doesn't control USB, but the stated companies do, AND the technology is free for all computer and device vendors.</I></BLOCKQUOTE><P>USB is Intel's baby. Just look at USB2. Intel decides where USB is going. It's not under the control of a standards organization like IEEE. The inclusion of "PC industry leaders" in teh USB camp is meaningless window dressing since Intel is far and away the heavyweight among them. What Intel says (and does with its products) goes. Intel is careful to trump-up the "standard"-ness of USB and the fact that other companies are "involved", but the bottom line is that it's still under private control, no matter how open or free Intel decides to make it.<P>[This message has been edited by John (edited November 29, 1999).]
 

John

Ars Praefectus
3,788
Subscriptor++
Which information? That USB is under the control of private companies and not a standards body like IEEE, ISO, or even the IETF? That's pretty much general knowledge, isn't it? The stuff about Intel being in the driver's seat is readily apparent when you look at the list of companies involved (Compaq, NEC, etc.) and consider the fact that in any grouping of PC hardware companies, Intel is the top dog. (MS is on the list as well, but they're pretty hardware-agnostic these days...although I expect them to roll in IEEE-1394 support just to keep Intel honest--see also 3dnow, etc. :) <P>For more hints, wander around usb.org and look at the contact information for press releases, the list of companies on the documents, etc. Intel's not going to come right out and say that they're promoting USB because they have more control over it and more to gain from it than IEEE-1394, of course, so there's no "smoking gun." I'm just making conclusions based on what's in Intel's best interest.
 

John

Ars Praefectus
3,788
Subscriptor++
It doesn't matter who brought PCI to market and who controls it...or USB or AGP for that matter. What matters is that if Intel had not supported them, they would not be included in every new PC. And if that's not control, then I don't know what is.<P>In the case of USB, there are no technological reasons for Intel to push USB2 over IEEE-1394...but there are plenty of business reasons. A USB2 future for PC's is in Intel's best interest.
 
Status
Not open for further replies.