Internal Firewire Hard Drives?

[Xtremist]

I'm just wondering if there are any currently available or if they're going to start producing them in the short to mid-term future... I'd really like to implement some Internal Firewire drives when I build my next system...

 

[thomsbrain]

Um, Firewire is limited to 50 MB/s (400 Mb/s). Why would you want to do that? ATA is much faster.
edit: on a side note, I build a peer-to-peer network over firewire once for a client. That was fast.

 

[Xtremist]

I've found that you rarely, if ever, get the speeds ATA drives boast. Sure they have the "potential" to do that, but the simple fact is they don't, even on a completely defragged drive. From my understanding Firewire is just a better transport method, although I could obviously be wrong...

Either way that doesn't really answer my question, but thanks for the comment anyway.

 

[gsaldivar]

The current bottleneck in hard drives lies with the drives themselves, not in the choice of interface.

Having Firewire as a native drive interface would give you absolutely no increase in drive performance.

 

[EXman]

Yup I went from a 7200rpm year old maxtor to a X15 cheetah holy cow now that is a difference. but for the money a Quantum 10K III is barely noticably slower 4.5ms instead of the Cheetahs 3.9 or new ones 3.5ms also the Quantum's peak is at 54Mbs and has a 8mb buffer.

ATA/100 and ATA/133 are kinda a misnomer the IDE drives 99.9% of the time will not fill up an ata/66 much less the higher brothers I think it is more marketing and Hype than real life usable speed.

 

[Xtremist]

gsaldivar,

Agreed, however from my understanding the interfaces will play a slight role, at least in the method it accesses the drive and how data is moved between interfaces. And my impression was that Firewire does this in a 'better' fashion than an IDE bus.


If there's one thing I don't like about computer-oriented forums is that everyone has gotta ask "Why would you want to do that?" when they think their idea is better... Quite simply, I just want to know if or when internal firewire HDD's will be available! If for nothing else, my own curiosity. At any rate, I would appreciate if comments stayed "on topic" in this thread. The topic is if and when internal firewire drives will be available, and who is making them. The topic is not why would you ever want to use firewire as your primary drive interface. Cheers!

 

[corkyg]

Yeah . . . sometimes when you ask for apples you get offered oranges. The answer to your question is, a qualified "yes." I have Firewire in all three of my systems, and use it for external HDDs and CD/RW burners. It isideal for that. The connectors are shielded cables similar to USB but different. They carry power in the 6-wire cables and do not in the 4 wire cables.

I have seen no internal IDE drives modified to accept these cables, and the current 1394 PCI cards I have seen do not have ribbon cable ports.

Firewire external drives are normally regular IDE drives fitted in a case which has a conversion circuitry for the 1394 standard cables. You would have to extract such adaption kits from the external case and find a way to mount them inside your computer and then snake the external cable inside.

Firewire is a good solid transfer means, and it is derived from SCSI. Essentially it is the separation of what was done to SCSI 2 to make SCSI 3, and then isolated as a separate channel.

 

[Xtremist]

corky-g,

Thanks for the response, it's appreciated. It's interesting, because I've seen FW cards that have an internal port, so I assumed they had, at least intented, it to be used 'inside' your case as well as outside. I may end up going with a single SCSI internal drive to store OS and application data and have my 'personal' and game storage on external drives. Would make it easier to pick and choose what I want to take to lan parties at least 😉

At any rate, thanks again for the reply.

 

[gsaldivar]

My mistake, I read the first reply to your post and for some reason associated it with your original question.

I'm thinking we will probably see SerialATA before native Firewire implementation on drives themselves.

 

[CQuinn]

<< From my understanding Firewire is just a better transport method >>



Partially correct, Firewire makes for a better external transport method,
as compared to most other interfaces (Serial, Parallel, USB).

I think what you are referring to is that IEEE1394 adapters have their
own controller, and do some level of disconnect and command queing services
among (multiple) connected devices. As opposed to IDE, which is limited
in devices per channel, and how well data transfer can take place between
devices on the same channel.

I don't think "internal" Firewire drive will be made available as a retail
solution any time soon, as that technology is best marketed as an external
solution. It is considered more convienient for the customer to only
have to connect the drive to the firewire port and plug it in to a power
strip, than to have to crack the case open and find a spare bay to install
it in.
Having said that, there is no reason someone could not create an internal
drive that has a built on firewire adapter to either connect to a firewire
port outside the case, or hack an adapter card to provides an internal
port to run 1394 cable from.

But, the benefit of that would be negligible compared to the cost.
The advantages for firewire (compared to internal IDE) only apply
when dealing with multiple devices on the connection, and that only
as far as bandwidth permits.

A quick and dirty comparison of:

Single (1 firewire vs 1 IDE (ATA-100))
- no difference, on either side the drives are not fast enough to
overload the bus during a sustained transfer. Firewire might not
be able to keep up with burst tranfers from some of the latest IDE
and SCSI drives.

Double (2 Firewire vs 2 IDE (ATA-100 on the same channel))
- Tie
Firewire would have an advantage accessing each device individually,
or transferring data from one drive to the other, but would saturate
the bus trying to access both devices at once, if both devices are
hard drives of the same speed and type.
IDE would not be as efficient at using both drives together, but would
have more bandwidth available to handle data transfer by the system.

2nd Double (2 Firewire vs 2 IDE (ATA-100 on different channels))
- IDE, would be equal to Firewire in treating each drive
separately, and would be less restricted by overall bandwidth.

Triple+ (3 or more Firewire vs 3 or more IDE (ATA-100))
- Firewire (as long as the speed doesn't go past available bandwidth)
the Firewire drives/devices can transfer data between each other with
less system overhead. Unless the IDE is using an extra controller to
run its drives each on thier own channel, or using a RAID configuration
that augements performance.

 

[Xtremist]

gsaldivar,

It's no problem. I would like to see some Serial-ATA devices soon too. I've heard some are shooting for Summer 02, but that's probably when motherboards with S-ATA controllers will start trickling into the market, followed months later with actual hardware. Of course I don't pay enough attention to know the most current status, but that's what I had read last I checked... I've been toying with the idea of a Firewire-based solution primarily because it's available now, and my Duron 750 is getting old... not really slow by any means, but my P4 1.8 at work puts it to shame 🙂 I don't think I can hold off long enough for Serial-ATA to become readily available.

CQuinn,

Thanks for the time you put into your reply. I agree with all your conclusions on IEEE1394 vs. IDE drive performance. I guess the only other thing I'm wondering/hoping is if each port on a Firewire card is considered its' own bus... Somehow I doubt it, although that would possibly alleviate some of the bandwidth issues you've mentioned since most adaptors have 4 external connectors... Thanks again.