Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
This is not a "Pro" machine.
- Thread starter timehacker11
- Start date
- Sort by reaction score
5gbps (or 10 for Ethernet) is in the same ballpark. If TB can't do it without chips in the cable, then the engineers are incompetent idiots and expecting the customers to pay to overcome a spec that is defective by design is pure stupidity.
huh???? What do you base that on?
What is TB used for besides external hard drives and video (where display port does just as well)? Show me a link to an external hard drive that's fast enough to saturate a USB3 connection.
Even if you can, which I doubt, why in the world would anyone buy it? You have fast, fast SSD connected internally to your pci-e bus, and the only thing external is mass storage, and then only if there's a good reason you want external (and wanting a computer that looks like a trashcan is not a good reason to lose your internal HD's). 5-10 gigabytes per minute from a cheap external drive over a $5 USB3 cable makes TB look like the garbage it is. Even if TB were faster, which it isn't in real life, it still isn't worth the cost to overcome the defective design.
TB is a solution to a problem the idiots at apple created (by removing the internal peripherals), and then it's a defective design because they require electronics in the cable. And even if you were stupid enough to buy into a TB desktop, you have a desk strewn with ugly little boxes with unreliable cables that can come loose instead of one simple box that has everything inside.
So yes, anyone who thinks this atrocious piece of garbage that even looks like the garbage can it belongs in is a pro machine, is so high on Apple Kool-Aid that it's amazing they can tie their own shoes.
If you think we live in some kind of PCIe expansion card heaven now for the Mac Pro, you need to have a look at some of the threads on the front page of this forum.
Dude, the fact is the current expansion options for the Mac Pro suck balls. We've been putting up with a lack of options, bizarre compromises, over priced cards, lack of EFI, bad drivers, and more for years. Where have you been? The sad truth is that all this is not really surprising... the Mac Pro market is a niche within a niche. Vendors have no reason to chase it.
IMHO, TB is a light at the end of the tunnel for OS X pros... it's a PCIe expansion solution for everything from the Air to the Pro with the Mini, iMac and Macbook Pro in between. Hopefully we'll finally start seeing some expansion options that don't require us to jump through flaming hoops to get things working. This coming from a guy who's so frustrated with PCIe expansion options for my 2009 Mac Pro, I welcome this new direction with open arms.
I think it's nice looking, reminds me of a turbine. I think anyone who used a previous gen Mac Pro can use this one as their Pro machine. I can usually tie my own shoes.
Yes, and this light is a train.
You really think that TB will change anything? With TB you have the same Driver and EFI problems, AND an Interface which is used almost exclusively on the Mac (another nail in the coffin).
TB has 4 times more bandwidth, it's not in the same ball park, at all. Look at cables with more than one channel of 10GbE, they are active.
Face it, what ever we have today has not been great. Most vendors can't be bothered to port a PC Card to the Mac, and when they do, it costs double and usually has issues or compromises of some kind.
Is TB the answer? The holy grail? I don't know, but it can't be worse than what we're dealing with now. Just have a look at the front page of this forum to see the issues people are having to add a USB 3 card, or a bootable Sata3 SSD that can do more than 1GB/s. It's ridiculous.
You simply need to compare the size of the market for Mac Pro PCIe cards to the addressable market for all Macs with TB ports (which is what?, 4M per quarter for the last two generations of Macs?) to see where vendors are going to invest their time and money in driver development, hardware qualification, and support. Time will tell, but I think we'll see vastly more TB peripherals than we ever saw PCIe cards certified and supported on the Mac Pro.
Are you too young to know anything about the evolution of computers? When a speed goes up 2-4x it's not even worth talking about. USB2 was 50 times faster than 1, USB3 was 10 times faster than that, and people talked about it. Sata 1 to 2 to 3 was doubling and nobody even really paid attention. Ethernet has gone from 10mbps to 100 to 1000 and now to 10,000 in those steps. Any smaller stepsize is a bad joke. Just like multichannel is a bad joke.
5 to 10 to 20 are minor sub-steps hardly worth nothing in the product spec. The fact that they require expensive specialized proprietary hardware to make those minor sub-steps makes the technology DOA.
I now see this is well before your time, but go google microchannel architechture, rambus, 2.88 megabyte "extra density" floppy discs, or 1000 other technologies that were a tiny bit better and used tech that was grossly overpriced compared to the mainstream cheap stuff that was almost as good.
If you're willing to pay $150 more for a 10 gigabit drive interface than a 5 gigabit one, then good for you, a fool and their money are soon parted, but there are not enough fools to keep TB from the trash heap of history where it belongs.
----------
It's not possible to debate aesthetics. One person's Ferrari is another person's scrap heap. It's one of the things that keep life interesting.
But even for someone who thinks the machine is gorgeous, I can't imagine anyone who was willing to pay for a previous pro even considering buying this. Lack of internal expansion defeats the whole point of a pro.
If you like a mini or imac, you probably will like the new pro, but you're not going to pay the premium when a mini does what you want, so it doesn't matter how many people say the pro is great when they're not interested in buying one. If you want a high-end machine and are willing to pay the high price, then the new pro is a total failure. No drive bays, limited ram slots, it's simply useless to it's target audience.
--------------
so you're the deciding voice on what can or can't be talked about? as well as the decider on what i want?
i like your style man! i don't even have to think or say anything when you're around.. just like a TV
Well no reason if you don't actually look at what the cable is actually doing.
TB actually does 20Gb/s through the cable, not 10. It is the whole TB network bandwidth not just the "half" of the data running through the cable. I think it was/is prudent for Intel to market it as 10Gb/s because that sets the expectataions correctly. Apparently, the vast majority of people look at only half of the TB system to evaluate it. Most of this whole thread is filled with such non- holistic viewpoints.
So, good luck finding a > 10GbE copper cable with no transceivers. Also good luck passing the FCC Class B (household , non-server room, ) emissions standards too. Finally, when you find one of those that pass those two items throw in "power over Ethernet" capability too.
Competent enough to be able to count up the whole bandwidth traversing the cable. Have to get past basic addition to get to the engineering aspects of the problem.
[ Aside.... on a Xeon E5 system with a C600 chipset ... USB 3.0 takes up PCI-e v2 lanes ("CPU laneways") . So competent enough also to know what does and does not require PCI-e v2 lanes ]
Kool-Aid. LOL.. There are plenty of kool-aid sellers on both sides in this thread.
Last edited:
There are physical limitations to what is possible, they have been reached a long time ago for CPU clock speeds for example. Active cables or optical are increasingly a necessity to get faster data rates over cable, look at InfiniBand, active cables, multiple channels (just like PCIe).
It's also funny that you only mention costs when it helps your argument, 10GbE cables costs much more than TB.
http://www.newegg.com/Product/Produ...MATCH&Description=10GbE+cable&N=-1&isNodeId=1
USB 1.0 12Mb/s 1996
USB 2.0 280Mb/s 2000 ( 4 years and 23x )
USB 3.0 5000Mb/s 2008 ( 8 years and 17x )
Cheap not fast was USB 1.0 objective (for example IEEE 1394, FW400, appeared in 1995). The relatively low starting point isn't really a long term historical trend. That's why the jump to still less than FW400 four years later wasn't that hard. USB 3.0 is a bigger, more aggressive jump but it comes 8 years later. With a 18-24 month Moore's law cycle that like 4-5 cycles later. Errr duh there will be a leap. The 2009 Mac Pro beats the slop out of a 2001 PowerMac.
Fair enough, but you miss the point .
Apple is marketing TB as a mainstream solution .
The Ethernet cables that plug straight into my Macs are a couple bucks each .
Those are not 10GbE, whatever that is, but they come stock .
Which is the entire point .
You DO know that thunderbolt does not exactly ... what is the word ... work over long distances?
True.
But a 2012 MP doesn't run circles around a 2006 MP .
That was the last MP leap, not 2001 or '09 .
Nothing suggests the new MP will be a significant step forward yet, re. performance . Or anything else, apart from size .
I'm afraid it's you who miss the point. 10GbE is 10Gb/s Ethernet, the regular cat5 Ethernet cables for a couple of bucks are used at 1Gb/s.
Apparently there are optical Thunderbolt cables, which do work over long distances.
I'm not dissing Optical, but I've not yet found any of them "in the wild" yet for sale. Got a vendor source and retail price on them?
FWIW, what this tangent is really heading towards is that there can be competing technologies to afford similar (or same) capabilities. One of the factors that will invariably be part of any trade-off decision will be each options' costs.
-hh
No price but I found that Corning has them on their website.
http://www.corning.com/CableSystems/OpticalCablesbyCorning/products/thunderbolt.aspx
Completely not material and largely misdirection. I just subtracted 8 from the 2009 (last major shift.). If USB 3.0 had taken 10 years ( pragmatically it did.... hardly anything showed up in 2009) I would have compared the 2009 MP to a 1999 PowerMac.
This has nothing to do with the 2006 MP. Even doing the -5 of the USB 1.1-2.0 transition would have put things back into the PPC era from 2009. The 2005-2006 transition is Apples to Oranges inflection point in primary cause in performance shift.
Chuckle. Not significantly better at Running PPC (Rosetta ) or 32 bit Intel programs perhaps. But for anything that scales that is a joke.
Great news .
So could you eleborate, which performance gains can we expext from an OOTB MP 6.x . .
Never mind 32bit programs or Rosetta, it's gone anyways .
Compared to a 2008/9-'12 MP , with comparable Ram and decent SSds installed for system and current data .
Oh, and zero external stuff, pleace .
So you're willing to pay how much more for a 2x speed increase on a communications port?
----------
The whole point is it doesn't matter what the cable is doing at a low level. It's transporting data at 10-20 gigabit. If they need smarts in the cable to accomplish that and that makes the cable cost $50, then it's Intel's design flaw.
----------
USB 2 was 480Mb/s.....makes my numbers pretty close.
Nice try.
----------
Very closeminded. I remember when HDD's were a few gig and we'd reached the limits of what was possible with platter density. Metal is not at it's end yet, and optical is pretty cheap and only beginning to enter mainstream.
Yes, very funny. 10GbE optical is about $25. For copper, cat 6 is rated up to 10 gigabit and is about $5 for a cable.
http://deepsurplus.com/Network-Stru...g-MTRJ-LC-50-125-Multimode-Duplex-Fiber-Cable
https://en.wikipedia.org/wiki/Category_6_cable
well, i'm looking at a 200GB lacie firewire drive that i paid over $200 for some time ago..
i'm willing to bet i'll have no problem spending the same amount on 2TB storage via thunderbolt.
or does that not answer your question properly?