Slashdot Mirror

As another post says: 10GigE is here now. A 10GigE card from Intel is already ~ $2K cheaper than an IB HCA. Though IB and 10GigE switches are still roughly the same price, that won't last long as more vendors start implementing 10GigE.

Going forward, which do you think will win? (Hint: the one that's 10 gigabit ethernet...)

But note that 10GigE cards are now available, and will likely mostly-kill Infiniband and Fibre Channel - even if IB still has a few technical advantages, it just won't be able to compete with commoditised and IEEE-standardised vendor-neutral 10 Gigabit Ethernet.

Is there faster ethernet available that I've not heard about?

Well, there's 10 gigabit Ethernet, and Intel are now sampling a card that supports it.

Really, the thing that needs to be done is improve the *protocol* that the communciation rides upon. If you're using gigabit, usually the limiting factor is the processor- you max the processor (doing the overhead for TCP/IP) before you even come close to filling up the gigabit pipe. There are technologies like Infiniband which try to address that, but progress has been very slow.

There is a big push for TCP offload- to take the effort for TCP and put it into an ASIC- check out this paper. TCP offload is absolutely necessary if we want to go beyond 1 gigabit ethernet any time soon, because the rate that we can communicate is beating the rate of increase in processing. IIRC Moore's law is an 18 month doubling time, there is a similar law for communications speed that says it doubles at a 12 month rate.

In addition, the engineering hassles of making a device with so many paralell channels is pretty hard- When you start doing high speed communications, you have to take a *lot* of care to make sure that the signals get there as you want them. This is why Intel appears to be in the process of moving away from PCI (a paralell bus) and moving to PCI-Express/3GIO, a combination paralell/serial bus, with a variable number of serial channels. PCI-Express is designed to carry on 16 wires what PCI-X carries on 80+ wires. When you actually have to make hardware- this makes a *HUGE* amount of difference.

It should be obvious that to burry copper is completely obsolete.

Wrong! Copper is already strung around every city and home in America (probably a hefty portion of the world). And, there's a standard for gigabit over copper:

Deployment Guide
[PDF]

It's limited to 100 meters, but for communities, home networks and any switched network, I don't see a point in passing up what is already laid in the building. For future digs, they could go either way, and I'll agree fiber is the way to go. But let's not ditch copper just yet...it seems to have some usefulness left in it.

802.3ae, as the IEEE lovingly calls it, is backed by the 10GEA (10 Gbps Ethernet Alliance). The founding members of the 10GEA are small companies you might have heard of such as 3Com, Cisco, Intel, Nortel or Sun.

Ugh. While I was as pleased as anyone that we were able to extend the functionality of twisted-pair to 125Mhz(Cat 5e) and 250Mhz(Cat 6), it's a pity that this has been widely interpreted as a sign that twisted-pair copper will indefinitely keep pace with fiber.
This just isn't true, and the sorry fact is that although cat-6 is in the marketplace, it only doubles the bandwidth of cat-5e, and at a premium cost (with today's run rates). Even worse: the few remaining communication standards intiatives that were planning on using cat-6, like 1000BASE-TX (not 1000BASE-T!), have lost all momentum in the marketplace.
Simultaneously, dozens of new standards have been announced or delivered for fiber. AFAIK, there currently isn't much development behind any copper-based high bandwidth standards. Most of the new high-bandwidth technologies aren't even looking at copper. 10GB Ethernet, for example, is defined as a fiber-only technology. And there are dozens of new high-speed projects on fiber, from 10GE through 2G/10G Fibrechannel to DWDM.

Video over Cat5 is cheap

If you're talking about NTSC over twisted-pair, I'd recommend against it. I notice significant rollof of high frequencies with high-quality baluns on Cat 6. Cat 5 is much worse. Coax is still the cable of choice for video.
On the other hand, if you're talking about video over ethernet or some other form of (presumably MPEG) distribution, sure this will work great if you have ethernet-enabled video equipment. My TV isn't.

Hi definition uncompressed video is more than 100Mb/sec - but 1Gb/sec over copper is on the horizon.

Uncompressed HDTV is about 1.5Gbps. 1Gbps ethernet over copper has been shipping for over a year: 802.3ab

Bottom line: For my money, I'd install a duplex of multimode fiber to each room. It's not significantly more expensive than Cat-6, and it's guaranteed to have a longer life.