NVIDIA and friends working on alternate USB 3.0 spec, SiS joins in, Intel uninvited from everybody's birthday parties
By Paul Miller 
posted June 14th 2008 2:45PM
posted June 14th 2008 2:45PM
Remember middle school? These guys do. NVIDIA, AMD, VIA and now SiS (only two capital letters? Not trying hard enough) have all teamed up in a fight against Intel of truly pubescent proportions. Intel has denied accusations of hiding the USB 3.0 spec, since it's not their spec to hide, and claims it has no obligation to disclose its actual host controller specification before it's ready. This apparently has the other chip makers scrambling to make their own host controller, so they aren't beholden to Intel's schedule. That could cause problems for the end product -- if they don't build theirs exactly like Intel's, and with Intel's already being on the market by the time they're done, they'll have to return to the drawing board and possibly delay their release by nine months. They claim this could give Intel two years of zero competition in the USB 3.0 space, but Intel figures since it plans to release the spec for free, is investing heavily in its development, and isn't done yet anyways, it doesn't owe those companies a thing. This just gets better and better.
The USB 3.0 spec specifies what the data packets look like (both logically and electrically) and the algorithm for their sequencing (command packet, data packet, status packet; how error packets can be returned, etc.). (It also specifies the connectors, cables, hubs, and other stuff like that.)
The Host Controller spec specifies the "machinery" that's needed to implement the USB 3.0 spec on a PC. It describes the state machine that transfers all the packets in the correct way, and it also describes the PC interface to the machine.
Because it also specifies the PC interface, it means that any chip that implements the Host Controller spec properly can just use standard software drivers to connect to the OS.
Since USB is such a damn complicated spec, it means that the state machine to control it is a damn complicated beast, and it will take lots of development time and debugging to get it right.
Since there will undoubtedly be "holes" in the USB spec (points which were not made completely clear), it is possible for different companies to make different state machines that deal with the holes differently. This can result in incompatibilities, where a device might work with one controller but not with the other.
This is why it is important for host controller developers to talk to one another, such that any holes in the spec can all be filled in the same way, minimizing the risk of any incompatibilities.
Uh, does anybody really think Intel cares?? Maybe AMD will join Sis and Nvidia and then intel will shake in thier boots, ur, um, yeah right.
If Intel did all the work then they deserve something. Why don't they all just split the cost?
Since they're heading in the direction of using optical transmission anyways, why not skip USB 3.0 and solely work on an optical connection?
USB 3.0 has an standard wire and optical versions, didn't you see the demo from one of the tech shows on these very pages ?
USB is so passe. Everybody knows that i.Link/Firewire is technically much much better then USB. Its just a huge conspiracy between Intel and a few other top dogs to keep i.link out of the business.
i.link s3200 should be alot faster then USB 3.0. I hope i.link becomes the standard. USB is so proprietary. Its just like the Bluray vs HDDVD war. HDDVD is much better but still lost... the consumers should benefit the most but unfortunately its just a huge conspiracy.
Firewire is dead. There's no development on it right now. It's too bad, Firewire was a much better standard for many things.
HD-DVD wasn't better than BluRay, let alone much better.
You idiot, firewire is not dead, it's used vastly in audio and video production.
But I know I should take it easy, this section of engadget is full of idiots like you.
I'm not an idiot. I didn't say it wasn't still in use.
One of my coworkers co-invented Firewire.
Firewire is dead. No one is currently doing any further development on it. There's little chance you'll ever see FW1600, let alone FW3200 implemented by any company.
I like Firewire, and I insist my computers all have it. I still use a Firewire video digitizer. But I'm not kidding myself. Firewire is on the way out. And why would that be surprising when Firewire drives cost more than USB 2.0 drives. Firewire video digitizers cost twice what USB 2.0 ones do (and FW ones don't even have built-in MPEG2/4 encoders).
oh god I can't even begin with you.
Yeah, that's about what I expected.
well my grandma invented firewire and she sez that daisy-chaining, target-firewire OSX mode, and self-powered (bus powered) devices is teh rox0rz
I wish more people were like your grandma, if more people appreciated what Firewire delivers like your grandma does, then it'd still be the premiere external computer data connection.
HD DVD is better than Blu-ray? Says who? Certainly not the universal language of mathematics.
50 GB > 30 GB
54 mb/s > 36 mb/s
Firewire demonstrates the depressing truth that technical superiority doesn't always matter. USB is an awful standard for high speed access but as far as end users are concerned it's good enough and since it uses the same cables and connections as USB 1.1, it is far more convenient.
Firewire is to all intents and purposes dead for end users. It might find a niche to live on in, but its number is up. The same could happen to eSata too.
USB-ray ftw !
"USB-ray ftw" ftw
Good end : intel show unfinish usb3. and everyone happy.
Sad end : only intel chipset mainboard support usb3. everyone make usb4 that backward comparable to usb2 but not comparable to usb3
worst end: intel contracts with UB40 to do the theme song
Well when I first hooked up my friend up with mediacom, I found out that he had a broken ethernet port on his laptop, unrepairable. I know the cards don't cost that much but at the time he didn't have the money, so I decided to hook the modem up USB. The speeds are pretty decent. So there are reasons to continue and make progress on USB.
remember one time, at band camp, I hooked up my friend up with mediacom, I found out that he had a broken ethernet port on his laptop, unrepairable. I know the cards don't cost that much but at the time he didn't have the money, so I decided to hook the modem up USB. The speeds are pretty decent. So there are reasons to continue and make progress on USB
+1 for band camp.
A USB format war?
Oh woop de do... / sarcasm
wait... so if that was sarcastic you mean you _do_ think it's a big deal?
OH MY GOD INTEL JUST LET YOUR SPECS BE OPEN! Unless everyone else's alternates are compatible with USB 2.0, USB 3.0 will obviously win, so why is everyone trying?
What's up with Nvida all of a sudden? have they been eating rooster instead of chicken?
they want to open their can of whoop-ass
engadget as a site seems to back the little guy most of the time, exceot intel, theyre in intels back pocket it seems. intel should just let the specs be open, if youre ganna open them anyway just open them now and maybe youll get some help from other companies to make it better faster, thats how open source works
Why panic, no device manufacturer is going to risk incompatibility, this will cost Creative, Sandisk et al. as they will all have to support both HCI in all their devices, except Apple who probably just support Intel's design and tell everyone thats better because macs use it.
dude... they'll support hydrochloric acid?
Yes obviously hydrochloric acid and not host controller interface, ever heard of context or do you not read the articles ?
joke |jōk|
noun
a thing that someone says to cause amusement or laughter, esp. a story with a funny punchline : she was in a mood to tell jokes.
The HCI stuff only affects the host. So devices that work with one HCI aren't necessarily going to have problems with another.
The device interface for USB isn't standardized at all.
eSATA by the way does not transfer power, whereas USB does. But there is POeSATA (Power Over eSATA) in the works.
USB doesn't transfer enough power to be really useful.
Hundreds of USB devices beg to differ.
OHCI and UHCI all over again. Just like when USB first came along.
Great non compatible usb ports and devices, just what we need
http://blogs.intel.com/technology/2008/06/usb_30_for_the_masses_dispelli.php
No Intel isn’t holding back the specification, the whole point of Intel investing heavily (gazillions of dollars and bazillions of man hours) into creating this ‘Dummies Guide’ is to enable the industry to start building USB 3.0 into their silicon as soon as possible, so why would Intel purposefully delay?
--------------
So the "gazillions of dollars and bazillions of man hours" can be rewarded with chipsets Intel sell on the market for a couple years with no competition?
After all you like to stress how much it's all costing and like to stress the benefits:
"After all, the sooner USB 3.0 hits the market, the sooner all you readers will be flooding your devices and hard drives with insanely large files requiring masses amounts of computational resources, improving your lives, and making you pleased that you bought a quad-core processor."
and your "sharing ways" despite gazillions spent:
"Intel plans to make this spec available early in second half of 2008 with a no-royalty licensing obligation (Basically: free, gratis, unpaid, zero dollars, free of charge, at no cost, on the house). This isn’t only because we are just nice guys, but it is also because Intel has set the bar for technology leadership and industry stewardship. It is Intel’s stewardship that has lead to USB being the most successful interface in the history of computing. + We at Intel love it when available processor performance is used to the max.
-----------
Yes, one could argue USB transfers use "available processor performance to the max". Har!
Seriously, the blog states so much money and time spent and then inexplicably expects us to believe there's nothing in it for Intel.
Call me cynical but I think it's total BS. Not that I don't think they have every right to the money, but when you're trying to lie about it like that, it rubs me the wrong way. In all likelihood, it's all about being first-to-market for at least a year while others play catch-up.
Nvidia seems to be in constant fight with Intel lately.... [Atom / super Atom CPU fight/race].... now USB 3.0 specs...
.this reminds me of the insane race/fight between big brother G and M$....
......I'm here just for the fun.... -- think Mike Tyson vs. Holyfield...