Posted on 12/17/2007 10:24:45 AM PST by ShadowAce
The IEEE 1394 Trade Association has announced a new FireWire specification that the group claims is capable of delivering up to 3.2Gb per second of throughput. The new interface (officially known as S3200) is directly based upon the 1394b/FireWire 800 standard and uses the same physical connectors, arbitration, and protocols as its predecessor. In theory, this should allow vendors to roll out S3200-capable silicon in a very short amount of time.
(As a clarifying point, please note that while 1394a and 1394b are largely considered synonymous with FireWire 400 and FireWire 800, the terms 1394c and S3200 refer to two different specifications. S3200 is an extension and acceleration of current 1394b technology, while 1394c offers 1394b speed and capability over standard Cat 5e cable. To date, no implementations of 1394c have been seen in the wild.)
There's no word on when S3200 devices might hit the market, but the 1394 Trade Association expects the standard to be fully ratified by early February. That's well ahead of USB 3.0, the closest competition to the spec. At IDF this year, Intel's Pat Gelsinger forecast USB 3.0 ratification sometime within the first half of 2008. S3200 could conceivably hit the market faster depending on its ease of implementation, but whether or not earlier device availability will translate into improved consumer uptake is open for debate.
Maximum theoretical data throughput will definitely be a point of contention between the two standards. Intel has stated that it expects USB 3.0 to be 10 times as fast as USB 2.0, which would give it a 4.8Gbps transfer rate. In contrast, the current iteration of S3200 will top out at 3.2Gbps. It's impossible to predict how much the throughput difference between the two standards will impact real-world device performance, but it's definitely a marketing edge that USB 3.0 proponents will lean on.
Of course, FireWire—up to and including S3200—has always offered certain advantages that USB lacks. Not only is it markedly less CPU-intensive due to its peer-to-peer nature (USB is master/slave), but FireWire is capable of delivering more power over a single cable. FireWire also allows for cable runs of up to 100 meters; USB 2 allows for a mere fraction of this, though USB 3.0 should increase cable lengths considerably.
Intel has not shared much on the degree to which the USB 3.0 standard will address the cable length and power concerns, though in all fairness, it may not matter. FireWire has held a notable technical advantage for years and has still failed to supplant the existing USB standard. FireWire's critics point to the standard's proprietary nature and royalty-based revenue model as the reason for its failure to dethrone USB. Whatever the reason, the fact remains that outside of the Mac world, FireWire ports are typically available only via add-in cards or on certain higher-end motherboards.
FireWire isn't just going to go away, however—it's currently included on a number of set-top cable boxes and is deployed in certain military situations. But that doesn't mean a new specification will trigger a fresh wave of peripherals, either. The peripheral interconnect market is already crowded; USB 2 is already popular, and eSATA support is growing. Combine this with the inevitable swarm of USB 3.0 products, and S3200 may end up buried, save for its continued presence and popularity in the market niches where FireWire has already established itself.
Dear Intel,
Please get rid of the ridiculous “feature” where the standard’s bandwidth is divided by the number of ports. When I hook up an external hard drive and a wireless mouse dongle, I don’t want the hard drive going half-speed because of a friggin’ mouse.
Signed, me.
Given that Firewire 400 gives more throughput than USB 480, some 15-70% faster (depending on the test), I'm betting the performance between the two will be close. Then the question is what the other advantages are. USB has only one serious advantage -- it's less expensive.
A second advantage is the proliferation of existing products already using USB. For a manufacturer to upgrade to 3.0 from the current 2.0 spec would be cheaper than switching techs to firewire, I would think.
It's a new standard. They'll need new chips, drivers and connectors anyway.
For a manufacturer to upgrade to 3.0 from the current 2.0 spec would be cheaper than switching techs to firewire, I would think.
Now we're just back to cost again.
Intel doesn't need for USB to out-perform firewire -- it's already "good enough" for most applications, and as long as it stays pretty close to FW in speed, I think its position is fairly secure. But Firewire's not going anywhere, and the biggest reason is a niche market the author doesn't mention. It's how you get video out of a digital camcorder.
I live my Macs, but I have no illusions about Apples influence on standards -- Firewire won't go away, but that's due to Sony more than Apple.
All true. In the next couple years the question of whether to go USB, Firewire or eSATA is going to be very specific to the situation.
Personally I like Firewire for high-speed, mainly because it gives enough power to run a big hard drive and it daisy chains.
"(USB is master/slave), " uh-oh....
jeeeeeezuss christ...
My reasons exactly. I have a bus-powered Firewire hard drive that's awfully handy to use with the laptop. No need for a second power brick, and I can use it when the laptop's running on battery.
hey...don’t whack the messenger http://www.cnn.com/2003/TECH/ptech/11/26/master.term.reut/
You’re not going to believe this, but some years back some bureaucrat in California tried to tell all the state’s vendors to not use the terms “master” and “slave” when describing IT products. It was too demeaning, perpetuated a caste-based society apparently.
Anybody have a link to that old story?
I know that story was discussed here on FR for a time....
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.