external hard drive: firewire or usb2.0?

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

addragyn

Golden Member
Sep 21, 2000
1,198
0
0
I hate to break this argument up for just a question, but that combo enclosure that was pointed out...does anyone know who sells them? I really want the TT-AlumiU2F (USB/FireWire 2.5" bus-powered) but can't find it anywhere online.

Consider it broken. :cool: I'm not sure what specific enclosure you're talking about. If you're a fancy boy and into looks Googie makes very unique enclosures. I prefer the simple aluminum enclusres myself and bought a few of these awhile back. They're not so unique anymore. The newer version of the Cutie, the CutieDX looks pretty snazzy IMO. The Lacie DataBank looks like a silver brick. WiebeTech sells good stuff, their enclosures are rebranded and can usually be found elsewhere for less.

************************

Ok, now you're just acting like a tool. You're losing the argument, so you just try to insult me. I love it!

There's a contradiction in your statement... isn't there? If you're insulted then you're definitely not reading my links!


"IT IS NOT SEMANTICS. USB uses a star topology. FireWire can branch or star. IT IS A FACT, IT IS NOT DEBATEABLE. This is fundamental. The inability to understand this basic concept casts a shadow of doubt about every point you made. This is elementary."

I see that you have conveniently ignored the other half of my statement. You cannot chain with a DV camera (at least none of the ones I've seen). You cannot chain with the B&K receiver. That is a fact. Why do you feel that this is so important to the technological superiority argument?

You said the difference between a tree and a star was "semantics". I quote you to you, "Chain...branch...WTF is the difference? You're just arguing semantics now?" It is not a rhetorical nitpick. USB cannot branch. FireWire can.

I think it's safe to assume DV cams don't have two ports because there is no good reason for them to chain. Unlike my hard drive enclosures which do have dual 1394 ports. Where chaining is useful! Server boards frequently have dual NICS onboard. Desktop boards don't. It's simply devices being designed for their application.

Let's try a picture this time. Can USB do this? NO.



I don't have time to read 30+ pages of a spec just to figure out what this thing does. So I went here instead. The gist of the article:
1) The goal was to transfer large amounts of data into a PC
2) Firewire was chosen because at the time (late 99 I think), it was "the only existing protocol that was in the right ballpark bandwidth-wise", and it wasn't proprietary.


mLAN is used to connect many devices in a chain. Can USB do that? NO.

Your second point is telling. Why was USB 2.0 developed when FireWire had existed for so long? It took resources to engineer USB2. We as consumers are paying for it. Why is Intel making the consumer pay for an inferior bus when a superior one already existed? Couldn't they have passed the saving on or even better used those resources on other non-redundant technology. Maybe we could've had Hyperthreading earlier? Or maybe FireWire 800 could've been developed earlier. It would also mean no fragmentation among computer peripheral interface. USB is a great low speed, low power, low cost bus. FireWire is a great high speed bus. They complement each other very well.



Both of those statements are true for USB 2.0 as well.

They may well be. Are these three?
1. USB chain.
2. USB devices can operate without a root/host.
3. USB can do peer to peer transfers.



USB 2.0/firewire photo scanner (too early for detailed perf specs, since it was just announced). Link

Let me clarify my statement on scanners/printers: For $300 or less, you have more options on USB 2.0 than you do for firewire, and they offer the same level of performance.


Moving the goalposts. Again. You originally said "Both busses can push the data fast enough, so it's a silly argument." It's not all about how fast the bus is. It's also about how the bus goes that fast. Need to dig a little deeper Encyclopedia Brown.


I was highly disappointed by the firewire cluster links that I found. Most of them aren't using 1394 for the interconnect, just for external storage. Here's what I was expecting.

Your link goes to a paid article. Read it free here. Can USB cluster like that? NO. Here's a FireWire storage article @ Oracle. Can USB do shared storage? NO. This is FireWire doing things USB can't because of FireWire's technological superiority.


If you don't think I'm reading the links you're posting, you just aren't reading my messages. I did a page of rebuttal on the tech brief. I told you that IFE doesn't seem to be in use anywhere. I looked at the B&K pic to show you it didn't supply power and couldn't be used to chain.

You still misunderstand and protest on *basic points* of contrast between USB and FireWire.


You haven't learned anything. I'm beating you over the head with the same incredibly obvious points, but nothing sinks in. You appear to back Apple's assertion that 1394 is technically superior to USB 2.0. I'm trying to compare the two to show that's not really the case. It is only natural to use terms like win and lose in that sort of a comparison. BTW, I used the term "win" exactly 4 times in about 8 pages of posts. I'm sorry if you think I'm harping on that point.

FireWire is technically superior to USB and USB2. Empirical evidence i.e. FireWire's dominace in high end applications and it's use in industry demonstrate. Comparitively by spec FireWire is also superior to USB. USB2 is a great low cost bus for consumer applications where *low cost* takes precedence. In markets where performance is a higher priority over low cost FireWire is being used. What do you know that all the engineers making these design decisions don't?


The latest USB article you pointed to is ancient, and provides no info on USB 2.0. But I'm already well versed in the technology, so it's not needed.

Who cares if it's old, the information is still correct. IMO You're not clueless but "well versed in the technology" seems optimistic.


"1.While touted as one of the great features of the technology, the reality is that not only do you need a PC in the mix.
- You were wrong, you've been shown multiple instances of FireWire being used without a PC. You may not
understand these uses but that doesn't make them any less real."
- I could use the same argument for you. I provided a list of scenarios that you don't understand.

Do it. Show me USB or USB2 being used w/o a PC. Just parroting my point back to me is hardly a rebuttal.


"2. What do you mean by "Firewire can do stuff USB can't in places USB can't go"?
- I have demostrated applications where FireWire is being used where USB could not. This a simple fact of USB limitations and FireWire's abilities."
No, you have provided a link to one application where 1394 COULD be used, if IFE was actually in existence. There's no reason why USB 2.0 couldn't be used as the connection interface on the other devices listed.

Yes there is. It's very basic and I've explained it 3 times now. Again; chain v. branch, star v. branch, host v. peer.


Quote: "3. Someone else on the thread says that firewire "murders" USB 2.0 when it comes to performance. That person clearly hasn't used both types of devices on a regular basis.
- While murder is definitely not objective, the meaing is clear. FireWire is faster than USB2. Collective real world experience in this thread agress with that. So do *all* the reviews linked. "
No. "Murders" in this context implies that there is a large performance difference. That's not the case. The performance is on par. I went to extremetech and searched for more USB vs 1394 articles. Words like "slightly faster" and "similar CPU utilization" were common in those articles. example

Why would you link to such a poorly done review? They don't even specify the file sizes being used! PCMark is a junk benchmark(I presume anybody who reads Anandtech would know that). I have as much regard for it as SiSoft Sandra. Search on the forum, I'm sure you'll find my opinion is corroborated. It's widely know the Oxford 911 is the best 1394 - IDE bridge. They didn't use it.

It's not hard to find reviews that say what you want.

Despite the speed advantage USB 2.0 offers over the old v1.1 standard, FireWire remains the performance king of plug-and-play connections. (TechTV)

Conclusion: FireWire Gives the Best Results (Tom's Hardware)

And though we tested the drive with its USB 2.0 connection, this chart shows only the MicroGB's FireWire results; the drive performed about 15 to 25 percent slower using its USB 2.0 connection. (cnet)

Though USB 2.0 is rated at a higher throughput speed, FireWire delivered faster performance on external hard drives when connected to a desktop. (PC Mag)



Maybe now you've learned something.

You're a fan of the former Iraqi Defense Minister? :Q

FireWire cannot chain. It is a Hollywood movie.
USB2 is faster. The Apple infidels lie.
The heathens speak of bus power.
There is no bus power.
It is non-believer semantics.


*This discussion has become circular. It is no longer interesting or useful. I don't see it presenting any new information to the forum.*


************************


For anybody else reading this thread PC Mag has some good simple & short pages on both USB2 and 1394. The pages are very short and definitely worth a few minutes if you're interested.

FireWire
Introduction
History
Pros and Cons
The Future
Fast Facts

USB 2.0
Introduction
History
Pros and Cons
The Future
Fast Facts

Enjoy!
 

RightHere

Banned
Jul 6, 2000
191
0
0
"There's a contradiction in your statement... isn't there? If you're insulted then you're definitely not reading my links!"

No contradiction. You are using the bold text to highlight points that you think I'm not getting. You make it sound like you have to hold my poor little hand on everything. You resort to that because I've pointed out the flaws in your assertions.

ENOUGH with chain vs. star already! USB has hubs. You plug a connector into a socket. I do not understand why you are fixated on the location of that socket.

Don't give me this peer vs. master/slave junk either. I've provided numerous examples where the devices do not talk to each other when connected together.

"I think it's safe to assume DV cams don't have two ports because there is no good reason for them to chain. Unlike my hard drive enclosures which do have dual 1394 ports. Where chaining is useful! Server boards frequently have dual NICS onboard. Desktop boards don't. It's simply devices being designed for their application."

That first statement is nothing but a cop out. Just because you can't envision a need for it doesn't mean that one does not exist. You stated that the devices can chain. The DV camera and the receiver can't chain. Even if you can't envision a chaining scenario for the camera, surely you can for the receiver.

"Your second point is telling. Why was USB 2.0 developed when FireWire had existed for so long? It took resources to engineer USB2. We as consumers are paying for it. Why is Intel making the consumer pay for an inferior bus when a superior one already existed? Couldn't they have passed the saving on or even better used those resources on other non-redundant technology. Maybe we could've had Hyperthreading earlier? Or maybe FireWire 800 could've been developed earlier."

I couldn't tell you why USB 2.0 was developed. It wasn't until recently though (last year, maybe two I think) where Apple dropped the "tax" they had on 1394. It wasn't very much (.50 or something), but in the quantities that OEM's ship in it tended to add up. That hurt the initial adoption of 1394. Just calling it an inferior bus doesn't magically make it true. You mention the consumer. That's where I'm coming from with my arguments. What are the advantages of 1394 over USB 2.0 for Joe Consumer?

Why do you refer to USB 2.0 as "redundant technology"?

I doubt that the people that work on USB 2.0 have anything to do with the HT stuff. Why do you feel that 1394b would've happened sooner if it weren't for USB 2.0? Do you think that it needed Intel or other big PC companies to put the pressure on the 1394ta to crank it out? (That's not a slap in the face to Apple...they just don't do the same kind of volume). Those speeds were talked about for a LONG time...definitely before USB 2.0 was finalized. What took them so long? I didn't see a lot of press about 1394b when it was announced at one of the recent Mac shows. I was expecting it to be a bigger deal. I think the only devices they demoed were a host controller and two storage devices (and I'd bet they both used the same chipset). In this biggerbetterfastermore world we live in, I would've thought that vendors would've jumped on it faster. Has someone like Via or Sis said they are going to integrate this into their chipsets yet?

"It would also mean no fragmentation among computer peripheral interface. USB is a great low speed, low power, low cost bus. FireWire is a great high speed bus. They complement each other very well."

Holy crap. Finally something we can agree on! :) I'd only add that USB 2.0 is a great high speed, low power, low cost bus.

"Your link goes to a paid article. Read it free here. Can USB cluster like that? NO. Here's a FireWire storage article @ Oracle. Can USB do shared storage? NO. This is FireWire doing things USB can't because of FireWire's technological superiority."
Thanks for tracking that down. I never said USB could do this, so relax. I read that Oracle article. That's exactly what I was talking about when I said I was disappointed. When you said firewire cluster, I assumed it was the interconnect.

"You still misunderstand and protest on *basic points* of contrast between USB and FireWire."
No I don't. I disagree with you, or I don't understand marketing technobabble. I understand the basic (and advanced) differences between the two technologies.

"FireWire is technically superior to USB and USB2. Empirical evidence i.e. FireWire's dominace in high end applications and it's use in industry demonstrate. Comparitively by spec FireWire is also superior to USB. USB2 is a great low cost bus for consumer applications where *low cost* takes precedence. In markets where performance is a higher priority over low cost FireWire is being used. "

Again...just because you keep saying it, it doesn't magically become true. "Dominace in high end applications" is really being liberal with the term dominance. Please detail the specific applications and/or industries where you allege that 1394 is dominant.

"What do you know that all the engineers making these design decisions don't?"
If you're talking about 1394, that's not a lot of decisions ;)

"Who cares if it's old, the information is still correct. IMO You're not clueless but "well versed in the technology" seems optimistic."
No, it's not. It doesn't include any reference to USB 2.0. Well versed in both technologies actually.

"Why would you link to such a poorly done review? They don't even specify the file sizes being used! PCMark is a junk benchmark(I presume anybody who reads Anandtech would know that). I have as much regard for it as SiSoft Sandra. Search on the forum, I'm sure you'll find my opinion is corroborated. It's widely know the Oxford 911 is the best 1394 - IDE bridge. They didn't use it. "

Sigh... I was adding additional "real world experience". You and others on the thread wanted some sort of proof, so I provided it. There are several other articles on that site and others (you can google too) that show the same thing I told you I saw: USB 2.0 throughput is faster than 1394 throughput in some cases. They clearly listed the testbed setup. I don't claim to know anything about benchmarking tools. You wanted numbers, you have them. Don't argue with me about them. I thought that Oxford's new bridge (922?) was supposed to be even better...isn't it? Few articles mention the chipsets in the drives, so it's impossible to determine which is the fastest USB 2.0 chip right now (it doesn't appear to be the first generation ISD chip though).

techtv link, cnet link, pc mag link - no mention in these links of technical superiority. Could just be hardware or software that needs to be tweaked still.

tom's hw link - shouldn't you rip the lack of testbed detail here? I guess not when the data works to your advantage. (BTW, I have the same feelings for Tom's site that you have for SiSoft)

Another cnet link for you - "Despite its slower 5,400rpm hard-drive mechanism, the Maxtor Personal Storage 3000LE handily outperformed the 3000DV FireWire (IEEE 1394) drive, thanks largely to the higher data-transfer capacity of USB 2.0."

And from your pcmag link -
"And since nearly all new mainstream desktop PCs have USB 2.0 ports, USB hard drives will be the market winners."
"For CD-RWs, test results were a mixed bag, though USB 2.0 showed a slight edge. "
"Don't feel sorry for FireWire. It may be rated 80 Mbps slower than USB 2.0 and built into far fewer PCs and peripherals, but it outperforms USB 2.0 slightly in many instances"

No "murder". No "flipping USB like a pancake". They are close in performance.

"You're a fan of the former Iraqi Defense Minister? "
Heard he's getting a job with the 1394TA real soon.
 

alexruiz

Platinum Member
Sep 21, 2001
2,836
556
126
Originally posted by: RightHere
Originally posted by: alexruiz
I posted that one, and I use both interfaces everyday..... in fact, I challenge you to SHOW me data of a HARD DRIVE that gets HIGHER performance on USB 2.0 over firewire...... There are none so far.

The supposedely "best" IDE-serial bridge for USB 2.0 (In system design) is nowhere as good as the Oxford 911. Do you have DATA to REFUTE my claim???
The only data point I have is based on visual observation of the test data. I didn't save a screenshot or anything. I don't even remember the specific tool used to perform the benchmarks (I wasn't running the PC). But I think I remember the device that was used: Iomega Portable Disk Drive. You can get a USB 2.0 or Firewire interface for it. As I remember it, the USB 2.0 interface was actually a hair faster (I think it was maybe a 1% difference). To be clear on my point, I said that 1394 did not murder USB 2.0 when it came to performance. They can be pretty close.

The ISD bridge is a first gen bridge. Surely there are faster ones out there by now.

What is the IDE bridge in that drive for firewire??? If it is not the Oxford 911, we are wasting our time
What are the spec of the internal drive??? If that is a 4200 rpm drive, wse are also wasting our time.

If the hard drive is SLOW, USB 2.0 can keep up with firewire. Get a better drive and you will see the murdering..... If your enclosure is 2.5", the Travelstar 40gnx is the only 2.5 drive that will benefit form firewire. If your drive is 3.5", a decent WD special edition on firewire will MURDER anything else on USB 2.0.....

Some persons here have reported getting as high as 35 MB/sec with WD special edition drives on firewire...... nobody has seen USB 2.0 tranfers over 22 MB/sec

My enclosure is made by simpletech (I bought the complete drive). ISDN 300 and Oxford 911, 2.5" With the included drive (Travelstar 40GN, 20 GB) the maximum transfer on each interface was 14 MB/sec, firewire having higher sustained rate. Swapping the drive for a travelstar 40gnx (5400 rpm, 8MB cache) I get as high as 25 MB/sec on firewire, but NEVER,ever above 22 MB on USb 2.0....... It can have sustained transfer of 24 MB/sec on firewire, but never above 20 MB/sec on USb 2.0.......

The original question of the thread was "What is better for a hard drive" so unless you have DATA showing advantage for USB 2.0 you have been debating about a proven point.

Alex

PS. Even myself (with English as second language) think that chained (linear - sequential) and branched (star) are absolutely different concepts.... so you should also review your spatial conception.
 

RightHere

Banned
Jul 6, 2000
191
0
0
"What is the IDE bridge in that drive for firewire??? If it is not the Oxford 911, we are wasting our time. What are the spec of the internal drive??? If that is a 4200 rpm drive, wse are also wasting our time."

In one of the examples posted, the chip was the NEC chip for 1394. What shall we compare it to on the USB side? Impossible to say, since few of the reviews bother to mention that. You asked me to show you specs of a USB 2.0 drive outperforming a 1394 drive. I showed it to you. SAME physical disk, different interfaces.

"If the hard drive is SLOW, USB 2.0 can keep up with firewire. Get a better drive and you will see the murdering..... If your enclosure is 2.5", the Travelstar 40gnx is the only 2.5 drive that will benefit form firewire. If your drive is 3.5", a decent WD special edition on firewire will MURDER anything else on USB 2.0..... "

Keep up is one thing. Beat is another. USB 2.0 beat 1394 in the example(s) posted. Extremetech is down right now, but I believe it was a 5400RPM 3.5" WD drive.

"Some persons here have reported getting as high as 35 MB/sec with WD special edition drives on firewire...... nobody has seen USB 2.0 tranfers over 22 MB/sec"

Yes they have. I linked it earlier, but here it is again.

"My enclosure is made by simpletech (I bought the complete drive). ISDN 300 and Oxford 911, 2.5" With the included drive (Travelstar 40GN, 20 GB) the maximum transfer on each interface was 14 MB/sec, firewire having higher sustained rate. Swapping the drive for a travelstar 40gnx (5400 rpm, 8MB cache) I get as high as 25 MB/sec on firewire, but NEVER,ever above 22 MB on USb 2.0....... It can have sustained transfer of 24 MB/sec on firewire, but never above 20 MB/sec on USb 2.0....... "

There are a lot of variables here:
1) Host controller
2) Bridge chipset
3) Driver quality

Impossible to say what's holding back the performance of this drive without extensive testing. You seem to think the ISD chip is the fastest. It was the first. Typically the first generation of any technology isn't going to be the best example of what's to come. IIRC, Oxford had a slower chip the first time out.

"The original question of the thread was "What is better for a hard drive" so unless you have DATA showing advantage for USB 2.0 you have been debating about a proven point."

Yes, it's a proven point. USB 2.0 is faster than firewire in some cases. Firewire is faster than USB 2.0 in some cases. There's no arguing this. The point is the performance is very close. It should not be used as the deciding factor when making a purchasing decision.

The ability to use a USB 2.0 device in a USB 1.1 port (though at a slower speed) gives you a backwards compatibility advantage that a 1394 device just can't touch. It is very very difficult to find a machine today that ships without a USB port. A large percentage of those machines include USB 2.0 ports. It is incredibly easy to find a machine today that ships without a 1394 port.

"PS. Even myself (with English as second language) think that chained (linear - sequential) and branched (star) are absolutely different concepts.... so you should also review your spatial conception. "

Nice attempt at an insult, but clearly you have misread my posts. I NEVER said that I couldn't tell the difference here. It's just that it doesn't matter. You plug a device into a socket. I tried to clarify that in my latest point (since the other guy isn't getting it either). I also pointed out numerous 1394 devices that do not allow you to chain.

Let me provide a better real-world example:

1394 host controller
<port 1> DV camera
<port 2> B&K Receiver
<port 3> iPod

(Grrrr...formatting is getting lost, so this isn't as easy to see anymore. Edited for clarity)

USB host controller
<port 1> Microsoft Game Voice
<port 2> Microsoft Natural Keyboard Pro -> <port 1> Mouse + <port 2> Nokia monitor/hub -> <port 1> Scanner + <port 2> Sandisk CF reader + <port 3> USB printer

Now...which one appears to be a branch, and which one appears to be a chain? See...it doesn't make a lick of difference. Joe consumer doesn't care. There's no advantage.
 

Eug

Lifer
Mar 11, 2000
24,165
1,809
126
While touted as one of the great features of the technology, the reality is that not only do you need a PC in the mix
Nope. You can plug a Firewire DV camcorder into a standalone DVD recorder and it will work just fine.

But most important to me is power. Some of the higher power laptop drives are kinda iffy running off of USB power, and the PS2 power adapter is not really a direct solution. (Not all computers have PS2 ports, and even if they did, sometimes they're not next to the USB ports.)

6-pin Firewire can power basically any laptop drive. Mind you the PC laptop designers out there seem to forget the importance of Firewire power, and design the PC laptops' Firewire without any power at all.

I refuse to buy a laptop without powered Firewire. I'd hate to have to carry around extra power cables for my 60 GB external 2.5" Firewire drive.

Yes, it's a proven point. USB 2.0 is faster than firewire in some cases. Firewire is faster than USB 2.0 in some cases. There's no arguing this. The point is the performance is very close. It should not be used as the deciding factor when making a purchasing decision.
Firewire 400 is faster than USB 2.0. Believe it. Current USB 2.0 chipsets are simply not up to snuff. Countless reports have shown the same. The thing is though, too many people buy slower non-Oxford 911 chipset enclosures, which are similar in speed to USB 2.0. Ie. if you compare slow Firewire chipsets to fast USB 2.0 then the speeds are similar. However, if you compare the fastest Firewire chipsets to USB 2.0, USB 2.0 will get smoked every time (assuming that the drive can keep up).

Impossible to say what's holding back the performance of this drive without extensive testing.
It's a laptop drive. Even with Firewire 800 it would not get too much past 25 MB/s. With desktop drives, as others have said, Oxford 911 Firewire 400 blows USB 2.0 away.

the Travelstar 40gnx is the only 2.5 drive that will benefit form firewire
Incorrect actually. Even my Toshiba 4200 rpm Firewire 2.5" drive will hit 25 MB/s. Fujitsu 4200 rpm drives are similar. There is also a Toshiba 5400 rpm drive that's pretty fast. Remember, the 5400 rpm Travelstar 40GNX's forte is not sustained transfer rate. Its sustained transfer rate is very good, but nothing to write home about. What makes it fast is the 8 MB cache and the lower latency.
 

RightHere

Banned
Jul 6, 2000
191
0
0
"Nope. You can plug a Firewire DV camcorder into a standalone DVD recorder and it will work just fine."

One example. I provided several other examples where you cannot do this with 2 firewire devices.

"But most important to me is power."
<snip>
"Mind you the PC laptop designers out there seem to forget the importance of Firewire power, and design the
PC laptops' Firewire without any power at all."

Glad you noticed that. I can understand this for laptops, since they want to try to conserve battery life. But there's no good explanation for why that B&K receiver linked earlier and the Mitsubishi Promise Module use the unpowered connector.

"Firewire 400 is faster than USB 2.0. Believe it. "

In some cases, yes. But I provided numbers that showed USB 2.0 beating 1394 in some cases. Please go read that info.

"Current USB 2.0 chipsets are simply not up to snuff. Countless reports have shown the same. The thing is though, too many people buy slower non-Oxford 911 chipset enclosures, which are similar in speed to USB 2.0. Ie. if you compare slow Firewire chipsets to fast USB 2.0 then the speeds are similar. However, if you compare the fastest Firewire chipsets to USB 2.0, USB 2.0 will get smoked every time (assuming that the drive can keep up)."

They aren't as fast as they can be, but clearly they are "up to snuff" if they beat 1394 in some cases. You hit on a point I've raised several times in this thread. We don't know what the fastest USB 2.0 bridge is and even if we did, that info is rarely mentioned in the reviews. The link I posted on the usb.org site showed 34 MB/sec, which is very close to the 35 MB/sec someone quoted on this thread for the Oxford 911. That article is more than 6 months old, so there may be faster chipsets out there by now.

"Oxford 911 Firewire 400 blows USB 2.0 away."

Again, look at the links I've posted. That is not an accurate statement.
 

Eug

Lifer
Mar 11, 2000
24,165
1,809
126
Glad you noticed that. I can understand this for laptops, since they want to try to conserve battery life.
My laptop provides full Firewire power. Like I said, I won't buy a laptop without this feature. USB 2.0 isn't a viable solution since it doesn't provide sufficient power to run a higher power laptop drive.

The link I posted on the usb.org site showed 34 MB/sec,
That's a PR blurb, not an independent test. I can also post Firewire numbers showing peak speed over 40 MB/s, but it's not real life.

But I provided numbers that showed USB 2.0 beating 1394 in some cases. Please go read that info.
As I've said before, every bench out there has the fastest USB 2.0 benches as being far slower than Firewire 400 Oxford 911, if a fast drive is used. There is NO USB 2.0 test to date that even comes close to hitting the max speeds Firewire 400 can attain.

Where's the advantage to an end user?
Huge advantage for a laptop user. No need to carry around hubs. For a desktop it's not as big of a deal since 1) you usually have more ports on the computer and 2) you can just buy a hub and leave it on your desk.
 

RightHere

Banned
Jul 6, 2000
191
0
0
"USB 2.0 isn't a viable solution since it doesn't provide sufficient power to run a higher power laptop drive."
So use a low power drive instead. It CAN be done w/ just bus power.

If the drive requires too much juice, there's a standard USB connection that can provide additional power. Don't know if there are any machines that use this port yet.

"That's a PR blurb, not an independent test. I can also post Firewire numbers showing peak speed over 40 MB/s, but it's not real life."

Oh, if you want to talk "real life", the minimal performance differences I've pointed out wouldn't be noticed in most cases.

"As I've said before, every bench out there has the fastest USB 2.0 benches as being far slower than Firewire 400 Oxford 911, if a fast drive is used. There is NO USB 2.0 test to date that even comes close to hitting the max speeds Firewire 400 can attain."

We seem to agree that the comparison of the fastest <whatever> bus device to the slowest <whatever> bus device is a useless comparison. You need to compare the performance of the individual USB 2.0 bridges (ISD, NEC?, etc.) to the performance of the Oxford bridge. Unless you have more specific data on the performance of these chipsets than has previously been supplied in this thread, you have no facts to back up your statement. 34 MB/sec "comes close".

BTW, it seems like there's a lot of overhead on 1394 too (only getting 70% of the max throughput).

"Huge advantage for a laptop user. No need to carry around hubs. For a desktop it's not as big of a deal since 1) you usually have more ports on the computer and 2) you can just buy a hub and leave it on your desk."

In the chain that I listed above (and we can remove the B&K receiver since you're talking about a portable scenario), how can you chain the ipod and the DV camera? You can't. And on most laptops, you won't even be able to connect both of those devices at the same time w/out a 1394 repeater in the mix. Doesn't sound like an advantage to me.

Just about every laptop has at least 2 and often 3 USB connectors built in. (And I'd argue that if you're going to lug around 3+ USB devices with your laptop, it probably won't be a big deal to grab a hub as well. There are hubs available that fit in your PCMCIA slot these days, and some that aren't much bigger than portable USB flash drives. So they are pretty compact.)
 

Eug

Lifer
Mar 11, 2000
24,165
1,809
126
34 MB/sec "comes close".
I task you to find a single real life benchmark showing that.

So use a low power drive instead. It CAN be done w/ just bus power.
Hmmmm... You seem to like press releases. Anyways, yes you CAN with a low power drive. Too bad that as drives get faster (ie. 5400 rpm) they use more power. There a number reports just in this forum indicating that their 2.5" 5400 rpm drives refuse to work with USB 2.0 without external power.
 

thorin

Diamond Member
Oct 9, 1999
7,573
0
0
Original brain fart by: RightHere
If the drive requires too much juice, there's a standard USB connection that can provide additional power. Don't know if there are any machines that use this port yet.
Uh if there AREN'T any machines using this connector/port then how can you CLAIM it's a "standard USB connection" ???????? :confused:

Thorin
 

RightHere

Banned
Jul 6, 2000
191
0
0
"I task you to find a single real life benchmark showing that."

32 MB/sec: link

"Hmmmm... You seem to like press releases. Anyways, yes you CAN with a low power drive. Too bad that as drives get faster (ie. 5400 rpm) they use more power. There a number reports just in this forum indicating that their 2.5" 5400 rpm drives refuse to work with USB 2.0 without external power. "

I pointed to a press release so you'd know that it's not someone just hacking their devices together to make it work. There are people designing devices that will run with only bus power.

I thought most 2.5" drives were 4200 RPM. I don't know what the difference in power is for a 4200 RPM drive vs. a 5400 RPM drive.
 

RightHere

Banned
Jul 6, 2000
191
0
0
Originally posted by: thorin
Uh if there AREN'T any machines using this connector/port then how can you CLAIM it's a "standard USB connection" ?
I said "Don't know if there are any machines that use this port yet". That does NOT mean that there aren't any machines out there that are using it. I don't claim to know every fact about every 1394 or USB device ever created.

When I said standard, I meant it more like industry standard or specification. It's not a proprietary connector.