USB 3.0 in new Mac Pro

[agtoau]

What are the chances that the new Mac Pros coming out this month or the next will be fitted with USB 3.0 ports?

[bozz2006]

0%

[Transporteur]

I think the chances are pretty good that the new Pro is equipped with both USB 3 and SATA 3. PC mainboards with these interfaces are available for a few weeks now.

[nanofrog]

I think the chances are pretty good that the new Pro is equipped with both USB 3 and SATA 3. PC mainboards with these interfaces are available for a few weeks now.
That doesn't mean we'll see it on the next MP's though, as the Gulftowns are still LGA1366, and use the same chipset as now (X58/5520).

For it to have 6Gb/s, it would need a different ICH attached to the chipset (ICH10 family is only good for 3.0Gb/s, and it can't even do that fully, as there's a limit of ~660MB/s across its DMI connection to the chipset), or use a controller that places the data on the PCIe bus directly (essentially a SATA III to PCIe bridge chip), and not use the ICH at all. But instead of it being a separate card you have to buy, its on the board.

USB 3.0 would be similar.

[camomac]

hopefully we will see blu-ray soon too..
but i think the odds are about the same.

[Cynicalone]

It took Apple a long time to move to USB 2.0 so I don't expect 3.0 for a while.

And then there is the news about Light Peak (http://en.wikipedia.org/wiki/Light_Peak). Which could have an impact on everything in the next few years.

[ncc1701d]

0%

+1

It took them a year to go 2TB HDD's. USB 3.0, no chance this round.

[JesterJJZ]

+1

It took them a year to go 2TB HDD's. USB 3.0, no chance this round.

They waited until 7200 RPM 2TB drives came out.

[VirtualRain]

It really has less to do with Apple and everything to do with Intel.

Unless Apple opts to add a discrete USB 3.0 chip (which is unlikely due to added board redesign costs), I don't think we will see it in the Mac Pro until Intel integrates it into the ICH logic. A few sources have indicated that Intel won't be integrating USB 3 until 2011...

http://www.techtree.com/India/News/No_USB_30_for_Intel_Chipsets_till_2011_Nvidia/551-107322-581.html

Honestly, I believe the most likely outcome for a 2010 Mac Pro is EXACTLY the same logic board and chassis with just Gulftown processor options.

[JoeG4]

NOTTACHANCE!

Also, 0% chance of the price dipping below 2 large.

I'd actually bet on them making a xeon iMac and calling it a Mac Pro. Just watch!

[VirtualRain]

I'd actually bet on them making a xeon iMac and calling it a Mac Pro. Just watch!

If they can have "Pro" laptops... why not? In fact, I'm surprised the quad core iMac's didn't get "Pro" appended to their name... iMac Pro.

[Dragonforce]

the new Mac Pros coming out this month

what ?

[pastrychef]

Does anyone know if there are any USB 3.0 PCI-e cards that work with Mac Pros? TIA

[300D]

Zero chance. USB is obsolete. Light Peak will be added well before USB3 and it will displace USB entirely within a few years.

[All Taken]

Zero chance. USB is obsolete. Light Peak will be added well before USB3 and it will displace USB entirely within a few years.


I wouldn't be so brash... I like facts - the fact is that intel showed off a prototype/makeshift Lightpeak Mac Pro motherboard last year.

This would appear to me that intel and apple are discussing implementation of lightpeak in mac pros as it was demoed on a mac pro board.

As for displacing USB, I doubt it - I feel it will take the same position as displayport tech, (Known but not widely adapted) at least for the first few years of it's life.

[nanofrog]

Zero chance. USB is obsolete. Light Peak will be added well before USB3 and it will displace USB entirely within a few years.
Light Peak will take time before it's adopted, and will depend on the cost of the transcievers. Most of the Light Peak parts out there are typically 180nm, and expensive. More than $50 per part right now (closer to ~$70 actually). There's new parts that are developed (65nm CMOS) that's intent is to lower the cost to ~$10 per part range. But so far, those parts are not widely available (evaluation components yes, but not so much on the supply side yet, as the production isn't to be ramped up until 2010). There's mulitple partners involved, and it all has to come together (different subcomponents from different suppliers).

USB 3.0 parts are already in the supply chain, and are backwards compatible with existing USB devices. So they're not going to suddenly disappear overnight. :rolleyes:

I wouldn't be so brash... I like facts - the fact is that intel showed off a prototype/makeshift Lightpeak Mac Pro motherboard last year.

This would appear to me that intel and apple are discussing implementation of lightpeak in mac pros as it was demoed on a mac pro board.

As for displacing USB, I doubt it - I feel it will take the same position as displayport tech, (Known but not widely adapted) at least for the first few years of it's life.
Light Peak is aimed as a replacement for a multitude of peripheral buses, such as drives (SCSI and SATA), external devices (FW & USB) and display (including audio signals) such as HDMI.

But it's not ready quite yet, as the parts aren't readily available in the supply chain, and it has to be cheap enough to attract adoption by device makers.

[goMac]

Light Peak will take time before it's adopted, and will depend on the cost of the transcievers. Most of the Light Peak parts out there are typically 180nm, and expensive. More than $50 per part right now (closer to ~$70 actually). There's new parts that are developed (65nm CMOS) that's intent is to lower the cost to ~$10 per part range. But so far, those parts are not widely available (evaluation components yes, but not so much on the supply side yet, as the production isn't to be ramped up until 2010). There's mulitple partners involved, and it all has to come together (different subcomponents from different suppliers).

USB 3.0 parts are already in the supply chain, and are backwards compatible with existing USB devices. So they're not going to suddenly disappear overnight. :rolleyes:

...except Light Peak supports attachment of USB 3.0 devices, letting you have your cake and eat it too.

[nanofrog]

...except Light Peak supports attachment of USB 3.0 devices, letting you have your cake and eat it too.
It's dependent on the inexpensive (smaller die) semi's, and I'm not so sure it will be manufactured in quantity in time for the next MP's. :(

[300D]

It was demod on a MP prototype logic board. I wouldn't be surprised if it was a prototype for the next MP release. Apple has been known to pick new technology for early release (USB 1.1, various wifi standards, Nehalem CPUs, PCI-express, etc).

[nanofrog]

It was demod on a MP prototype logic board. I wouldn't be surprised if it was a prototype for the next MP release. Apple has been known to pick new technology for early release (USB 1.1, various wifi standards, Nehalem CPUs, PCI-express, etc).
Actually, it wasn't. It was an Intel engineering system (i.e. Frankenstein system) that was hacked to run OS X. :eek:

Apple's involvement is a rumor, not a confirmed fact. Now assuming it is true, the article (Engadget (http://www.engadget.com/2009/09/26/exclusive-apple-dictated-light-peak-creation-to-intel-could-be/)), states that it was nothing more than a concept, not a full fledged design. Intel and a few other partners did all the actual work to bring it to fruition (the demonstration was a prototype system; evaluation board stuffed in an Intel system - note the perf board and hand written text on the board "Light Bridge SV Board, Rev. 1.0").

They're hoping it will debut in 2010, but it will depend on getting the semiconductor components completed in a die size that's inexpensive enough where board makers (and companies that would produce PCIe cards) will bite. If it's too expensive, they'll wait for cheaper parts, and Intel and their partners are well aware of this. That's why they're trying to hit a $10USD price point for the sytem (and it's more than just a chip).

[xgman]

Actually USB 3 will end up sooner than later on all new macs. Especially the Mac Pro would benefit from this fast interface for media transfers and storage. USB 3.0 is already on some new motherboards and there is an add on card although no mac drivers yet. I think usb 3.0 has the potential to be adopted quickly and has a hugh spped icrease immediate impact whereas sata 3 so far has shonw little or no real increase in speed.

[nanofrog]

Actually USB 3 will end up sooner than later on all new macs. Especially the Mac Pro would benefit from this fast interface for media transfers and storage. USB 3.0 is already on some new motherboards and there is an add on card although no mac drivers yet. I think usb 3.0 has the potential to be adopted quickly and has a hugh spped icrease immediate impact whereas sata 3 so far has shonw little or no real increase in speed.
Yep. USB 3.0 already has the parts in the supply chain, and they're inexpensive.

SATA III has potential, but will really only matter with SSD's, or mechanical on a PM enclosure (one eSATA port for multiple drives that can exceed the current throughput limit of ~250MB/s with existing SATA II PM chips) on the consumer/workstation front.

[300D]

Apple's involvement is a rumor, not a confirmed fact.
Ah, Intel just "chose" to demonstrate on OSX instead of the OS thats on 95% of the world's computers? :rolleyes:

[nanofrog]

Ah, Intel just "chose" to demonstrate on OSX instead of the OS thats on 95% of the world's computers? :rolleyes:
UNIX is easier to develop for when developing IC's, and there's more TCAD simulation software for UNIX than Windows.

It's also easier for testing IMO (i.e. VM). Get it working under UNIX, and you've proven the hardware works. Then tackle Windows. It narrows the issues to software by that point, and you're not pulling your hair out going at it from both directions.

[goMac]

UNIX is easier to develop for when developing IC's, and there's more TCAD simulation software for UNIX than Windows.

It's also easier for testing IMO (i.e. VM). Get it working under UNIX, and you've proven the hardware works. Then tackle Windows. It narrows the issues to software by that point, and you're not pulling your hair out going at it from both directions.

If they were running OS X, you can bet that Apple wouldn't have approved of them running OS X on non-Apple hardware.

OS X doesn't share a driver architecture with UNIX at all. "Standardized UNIX" doesn't really come into play at all. The OS X driver API is likely even more obscure than the Windows one.

[nanofrog]

If they were running OS X, you can bet that Apple wouldn't have approved of them running OS X on non-Apple hardware.

OS X doesn't share a driver architecture with UNIX at all. "Standardized UNIX" doesn't really come into play at all. The OS X driver API is likely even more obscure than the Windows one.
The history of Light Peak's origins are in dispute though, and what's interesting to me, is that Apple was NOT even mentioned at the IDF conference. It's not a secret project by any means, and it would be good PR for Apple to have their name associated with it had they been in it from the beginning.

That said, I do think Apple's highly interested, and should even be willing to spend some funds for code development. It's cheap, and less than peanuts when you think about what the other developers have spent on this.

A fast, optical 10Gb/s interface would be highly desirable, and notably in the enterprise market. If you think about it, it's as fast as 10G Ethernet without the additional overhead (IP stacks, routing,...). That gives a high speed interface directly in the system without the need for a card (though it does use PCIe lanes to connect to the system, they're routed directly on the board).

This is something Intel would be highly interested in, and the MP is NOT the only systems that use EFI. The Itaniums do as well (where EFI originated BTW), and why we actually have EFI based RAID cards. They only needed to develop drivers to enable it to function under OS X. ;)

As Intel's products benefit Apple (they wouldn't have any computers without them right now), they'd most likely not complain. They'll certainly benefit from Light Peak hitting the market as well. And even though OS X's drivers aren't as easy as other versions of UNIX, it's still not that hard to do. I do think it would have been easier than Windows (which is the biggest PITA out there to develop hardware on IMO).

Why do you think drivers are so buggy/temperamental, even after a products been around awhile? :eek: ;)

[TennisandMusic]

If they were running OS X, you can bet that Apple wouldn't have approved of them running OS X on non-Apple hardware.

OS X doesn't share a driver architecture with UNIX at all. "Standardized UNIX" doesn't really come into play at all. The OS X driver API is likely even more obscure than the Windows one.

Non-Apple hardware?

Intel MAKES the hardware for Apple, and Apple has it cheaply manufactured elsewhere. Apple doesn't make anything at all, except for the software. Apple only brands hardware.

[nanofrog]

Non-Apple hardware?

Intel MAKES the hardware for Apple, and Apple has it cheaply manufactured elsewhere. Apple doesn't make anything at all, except for the software. Apple only brands hardware.
Though Apple is a "hardware" company (where they derive their profits), they don't manufacture anything, nor produce finished designs. They create an industrial design (Ives' function) and a specification document, and have it ODM'ed by someone else. Unfortunately, that's the way of most products these days. :( (And why so much is crap). :mad:

[300D]

Apple doesn't make anything at all, except for the software. Apple only brands hardware.
...
Though Apple is a "hardware" company (where they derive their profits), they don't manufacture anything, nor produce finished designs.

Clearly both of you have zero understanding of Apple's R&D department.

[nanofrog]

Clearly both of you have zero understanding of Apple's R&D department.
Prototype /= Finished product

And even when they have their hands on a prototype, that does NOT mean they did the actual design work (circuit design, PCB layouts, part selections, cost analysis, ....). They need such systems to develop the code for it. :rolleyes:

[xgman]

CES has announced so many USB 3.0 products that I can't even keep track of them. USB 3.0 is coming fast. Apple should react.

[300D]

And even when they have their hands on a prototype, that does NOT mean they did the actual design work

The same is true for ALL computer makers. The computer as it exists today is because a bunch of people decided to copy IBM with off the shelf parts.

[nanofrog]

The same is true for ALL computer makers. The computer as it exists today is because a bunch of people decided to copy IBM with off the shelf parts.
It's for different reasons (remember, there's not much out there for CPU choices. ARM has more variants than most, and is too limited for desktops, workstations, or servers, so that leaves Intel or AMD right now. If ARM gets to the point it can compete, who knows... But in simple terms, the silicon is super expensive to develop, and most can't afford it. The most recent Fab facilities are over $3 Billion USD to build. :eek: That doesn't touch the R&D in both the chip design and materials aspects (manufacturing process).

1. Standardization
2. Costs. No company wants to maintain a bunch of engineers to take a product from concept (paper) to a finished state, and maybe manufacture it. The manufacturing facilities were the first to go, where it was then produced by OEM's. Then the bulk of the engineers (and all in some cases), and changed from OEM to ODM.

Basic design specs, and maybe, such as Apple, an industrial design is created (packaging). That information is given to the ODM/s, which then does all the circuit design/development and manufacturing work. Prototypes are sent back for code development and verification testing. They eventually reach a point where it's given the go ahead for full scale manufacturing.

[maskedscientist]

hopefully we will see blu-ray soon too..
but i think the odds are about the same.

Never gonna happen. The concept of optical media is already outdated. Downloads are the future, man. even if you can't get 1080p. Apple has already started eliminating optical drives, like in the Air.

[maskedscientist]

Apple was the first to brink firewire, usb 1.0 and 2.0, CD drive. THey were the first to get rid of the floppy, and are getting rid of firewire. We can expect usb and sata 3 in the next pros.

[Santoanderson]

Never gonna happen. The concept of optical media is already outdated. Downloads are the future, man. even if you can't get 1080p. Apple has already started eliminating optical drives, like in the Air.

I wish they'd keep it as a contigency, or at least make it a ridiculously expensive 3rd party option. I won't totally embrace abandonment of physical media until I can download the equivalent of a Blu Ray disc, and it's features, in the same amount of time that it takes me to drive down to Walmart and back.