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.
SuperSpeed USB 3.0 Promises 10Gbps Data Transfers, Coming in 2014
- Thread starter MacRumors
- Start date
- Sort by reaction score
License prices need to go down
I like thunderbolt a lot but its going to die quickly if Apple and Intel don't lower the license fees immediately. I don't know if Apple charges any money on the use of there thunderbolt connector if so they need to lower or better yet make it free. I think Intel as well has to understand that for thunderbolt to live they are going to have to reduce the license fees substantially.
Every time I hear a manufacturer say why they don't have thunderbolt peripherals they say it's because of high fees.
I like thunderbolt a lot but its going to die quickly if Apple and Intel don't lower the license fees immediately. I don't know if Apple charges any money on the use of there thunderbolt connector if so they need to lower or better yet make it free. I think Intel as well has to understand that for thunderbolt to live they are going to have to reduce the license fees substantially.
Every time I hear a manufacturer say why they don't have thunderbolt peripherals they say it's because of high fees.
MOTHERBOARDS:
Give me a 8 x 8GB (8 x 16GB) or more Memory slots for DRAM for Intel/AMD Motherboard design specs for general consumer usage.
This 32GB artificial ceiling will soon force an entire reboot of the industry.
The baseline of 8GB is the new norm, though it should be 16GB for general systems. 32GB is nice, but one can max that out very quickly. We live in a computing age where software design cares far less about application optimization and memory resources than they once did. Even amidst memory pool sharing designs you eventually hit a wall, and actually sooner rather than later.
COMPONENTS:
Require a much higher component standard for all PC Motherboard manufacturers to ramp up the quality of voltage regulation, reduced power heat dissipation and shielding from other components to produce better work all around from basic to professional audio, video and engineering application markets.
I could care less about USB 3.x. Thunderbolt should be an open standard on Intel, AMD and other 3rd party chipsets. USB is inferior to FireWire and inferior to Thunderbolt/LightPeak.
Stop wasting our time, Intel, AMD, etc., etc. The amount of years wasting on inefficient components costs the world tens of billions in US Dollars, annually.
I should be able to purchase a Thunderbolt Display for a Mac and other 3rd party systems. [VR-Zone says it's coming: http://vr-zone.com/articles/asus-s-...to-get-intel-thunderbolt-support-/16207.html]
That article is from June 2012.
This leads me to my next hope:
DISPLAY QUALITY AND SIZE OPTIONS:
The quality of displays has declined substantially by third parties over the past 24 months. The high quality monitors form EIZO are insulting on the pocket book.
HP continues to decline into oblivion with their cheap replacements to the LP line.
Apple has a 27" T-Bolt and the old 27" Cinema. Better than nothing, but come on. It's 2012. That's truly pathetic.
We want high quality, hi-Res displays. Not cheap crap. We want quality for a price that means quality and not another round of ``Panel companies LG, Samsung, etc., are once again being fined nearly $1 billion for colluding on panel price fixing....''
This is getting old. Another ASUS June Article about a Thunderbolt Display [http://www.engadget.com/2012/06/07/asus-shows-off-its-first-thunderbolt-monitor-along-with-3d-and/] It's cheap and of low quality. Guaranteed to cost too much.
And here we are discussing USB 3.0 future versions.
Truly sad on the ramp up to CES 2013.
----------
Peanuts to the cost of data processing and distribution of bandwidth.
Give me a 8 x 8GB (8 x 16GB) or more Memory slots for DRAM for Intel/AMD Motherboard design specs for general consumer usage.
This 32GB artificial ceiling will soon force an entire reboot of the industry.
The baseline of 8GB is the new norm, though it should be 16GB for general systems. 32GB is nice, but one can max that out very quickly. We live in a computing age where software design cares far less about application optimization and memory resources than they once did. Even amidst memory pool sharing designs you eventually hit a wall, and actually sooner rather than later.
COMPONENTS:
Require a much higher component standard for all PC Motherboard manufacturers to ramp up the quality of voltage regulation, reduced power heat dissipation and shielding from other components to produce better work all around from basic to professional audio, video and engineering application markets.
I could care less about USB 3.x. Thunderbolt should be an open standard on Intel, AMD and other 3rd party chipsets. USB is inferior to FireWire and inferior to Thunderbolt/LightPeak.
Stop wasting our time, Intel, AMD, etc., etc. The amount of years wasting on inefficient components costs the world tens of billions in US Dollars, annually.
I should be able to purchase a Thunderbolt Display for a Mac and other 3rd party systems. [VR-Zone says it's coming: http://vr-zone.com/articles/asus-s-...to-get-intel-thunderbolt-support-/16207.html]
That article is from June 2012.
This leads me to my next hope:
DISPLAY QUALITY AND SIZE OPTIONS:
The quality of displays has declined substantially by third parties over the past 24 months. The high quality monitors form EIZO are insulting on the pocket book.
HP continues to decline into oblivion with their cheap replacements to the LP line.
Apple has a 27" T-Bolt and the old 27" Cinema. Better than nothing, but come on. It's 2012. That's truly pathetic.
We want high quality, hi-Res displays. Not cheap crap. We want quality for a price that means quality and not another round of ``Panel companies LG, Samsung, etc., are once again being fined nearly $1 billion for colluding on panel price fixing....''
This is getting old. Another ASUS June Article about a Thunderbolt Display [http://www.engadget.com/2012/06/07/asus-shows-off-its-first-thunderbolt-monitor-along-with-3d-and/] It's cheap and of low quality. Guaranteed to cost too much.
And here we are discussing USB 3.0 future versions.
Truly sad on the ramp up to CES 2013.
----------
Peanuts to the cost of data processing and distribution of bandwidth.
Doesn't change the fact that the USB protocol overhead keeps being a huge nuisance while they keep trying different solution (UAS) current performance is around 60% of the interface maximum theoretical throughput, and that not even talking of the tax imposed on the CPU.
The fact is USB 3 will never manage TB real world max data rates no matter what, since its always being a cheap interface for cheap devices (Single HDD, mouse, keyboard etc.), and that's OK.
The fact is USB 3 will never manage TB real world max data rates no matter what, since its always being a cheap interface for cheap devices (Single HDD, mouse, keyboard etc.), and that's OK.
So does this mean that all USB3 SuperSpeed peripherals that are out today will get 10Gbps with new USB3 cables? Or do they need to release updated peripherals with updated USB3 connectors/chipset?
One can't simply compare USB3.0 with Thunderbolt like they are offering the same capabilities.
USB3.0 is purely for transfers, and it's decent enough for 90% of users scenarios as hard drive speeds are not currently shattering the max transfer speeds of USB3.0 and people tend to not really care too much about access time and CPU utilisation when it comes to external storage.
Thunderbolt on the other hand is something completely different, it provides direct access to the internal PCI-Express layers making it basically a docking port where you can dock anything from high-performance external graphics card, to regular docks providing addition external I/O capabilities and as it's based on the mini-displayport connector it also doubles as a displayconnector providing compatibility with everything from VGA, DVI to HDMI at up to 2560x1600@60Hz.
There is no overhead to think about, the CPU utilisation is at a minimum and with direct access to the PCI-Express layers you mostly get access times and speeds like your external devices wore connected internally to the PCI-Express connector.
USB3.0 is purely for transfers, and it's decent enough for 90% of users scenarios as hard drive speeds are not currently shattering the max transfer speeds of USB3.0 and people tend to not really care too much about access time and CPU utilisation when it comes to external storage.
Thunderbolt on the other hand is something completely different, it provides direct access to the internal PCI-Express layers making it basically a docking port where you can dock anything from high-performance external graphics card, to regular docks providing addition external I/O capabilities and as it's based on the mini-displayport connector it also doubles as a displayconnector providing compatibility with everything from VGA, DVI to HDMI at up to 2560x1600@60Hz.
There is no overhead to think about, the CPU utilisation is at a minimum and with direct access to the PCI-Express layers you mostly get access times and speeds like your external devices wore connected internally to the PCI-Express connector.
Well if this does become possible, then something kind of tells me that you won't really need anything other than USB
Doesn't this sound better than Thunderbolt? Backward compatible, widely adopted, Mac + PC, cheap cables and devices, probably will stay around for many years to come, etc
While it'll be good to see competition, I do have to agree with the sentiment that Thunderbolt is the better standard overall; mainly because it's practically just an external PCIe port, which makes it fast and efficient and potentially very powerful for just what it can be used to do for awkward to upgrade devices (which is actually most of Apple's lineup these days).
Still, it'll be good to see a faster USB standard as well, hopefully keeping it around for longer, as the last thing I want to do is start fiddling around with hubs to keep all my existing USB devices available.
Still, it'll be good to see a faster USB standard as well, hopefully keeping it around for longer, as the last thing I want to do is start fiddling around with hubs to keep all my existing USB devices available.
I've read this entire thread and I am appalled by the number of people who are so confusingly comparing USB and Thunderbolt.
Both technologies are not competing with each other. They are complementary! There will never be cheap Thunderbolt peripherals because that's not what the technology is aimed at. There will never be benchmark shattering high-end USB devices because that's not what the technology is aimed at.
Thunderbolt is a PCIe extender, basically, the best there is for things like DAS (much simpler than 10GbE iSCSI setups or pure FC SANs, if not as convenient on a provisionning basis), docking stations or desktop replacement boxes (the Sony PowerMedia Dock is exactly what Thunderbolt is designed for, too bad Sony didn't actually use Thunderbolt for it).
Monitors ? Pure DisplayPort is still more up-to-date than Thunderbolt unfortunately, that is something Intel really goofed up. Keeping up with DisplayPort specifications should be of the upmost importance because otherwise TB is only gimping monitor connections.
USB 3.0 however is the quite and dirty, cheap connectivity for everyday devices. Things like thumb drives, keyboards, mice, trackpads, printers, single spinning hard drives for backups, optical drives, etc.. No need for Thunderbolt there, the bandwidth just isn't required, nor are the more costly and complicated implementations.
Both technologies can co-exist, both have a purpose. Stop trying to make this out into a competition, it's not, just like your Mini-van isn't competing with a Ferrari 360 Modena. I wouldn't drag 3 kids to a pic nic in the Ferrari, and I wouldn't go for a quick ride alone in the twisties in the van.
Much as I like to joke that 2013 is the "13th annual Year that Linux on the Desktop Really Takes Off", I'm afraid I have to concede that the TB boosterism sounds like the "3rd annual year that Thunderbolt Really Takes Off".
Also, it's crazy to tout the "pro" uses of Thunderbolt when Apple still doesn't ship a Mac Pro with Thunderbolt ports. Yes, there's a Pro laptop with TB, yes some designers and video pros will make do with an iMac or a Mini, but it still can't help but symbolize the neglect of the high-end desktop Mac, the one for which you'd actually be able to sell $1,000 peripherals. The more I think about it, the more I'm resigned to the idea that TB devices can't come down in price if there's never a market for them in the first place.
Also, it's crazy to tout the "pro" uses of Thunderbolt when Apple still doesn't ship a Mac Pro with Thunderbolt ports. Yes, there's a Pro laptop with TB, yes some designers and video pros will make do with an iMac or a Mini, but it still can't help but symbolize the neglect of the high-end desktop Mac, the one for which you'd actually be able to sell $1,000 peripherals. The more I think about it, the more I'm resigned to the idea that TB devices can't come down in price if there's never a market for them in the first place.
The interesting part is that many PCs have HDMI connectors which take up more space, only do video and the cables cost the same amount. No one complains about that. Thunderbolt can support up to 7 devices including the video as well.
Uh ? HDMI cables are like 3-5$ a pop.
Unless you're talking about those scam artists that sell them for more than that ?
I agree 4K display is the primary end user use case for TB and where 10GigE (Cat6e) also suits the need. HDMI currently supports 4K. Apple persued Light Peak with Intel to get past the I/O bottleneck using a capacity that already exists in every board. PCIe. This is a facility that may be an element of the next MacPro to allow better graphics and display expandability, as well as AppleTV next to facilitate multi-room HD (1K) and main room 4K. The main benefit to the connection method is it will be adopted by a mass consumer manufacturer, so whether you need it or not it is readily available.
I'm not sure about the practical throughput of 802.11ac or ad but it is probably close too. At pretty short range like no more than 50-100 feet in real world installs.
On a side note, Sony is in the process of converting its more popular titles to 4K right now, so they know something. I doubt they plan theatrical re-releases. I saw Lincoln last month on a traditional theatrical screen in 4K. It was not pixelated.
I'm not sure about the practical throughput of 802.11ac or ad but it is probably close too. At pretty short range like no more than 50-100 feet in real world installs.
On a side note, Sony is in the process of converting its more popular titles to 4K right now, so they know something. I doubt they plan theatrical re-releases. I saw Lincoln last month on a traditional theatrical screen in 4K. It was not pixelated.
Last edited:
Exactly my point, we have yet to reach a critical mass, and that milestone seems a long way away still, which is why Intel should be subsidising the TB chip or building it into their platform hub.
So the Ars article was wrong, we've got another year to wait before the even the cables come down in price... Great. As another poster pointed out, 10Gb Ethernet runs over very ordinary CAT6e cable with less wires, so why couldn't Intel have done something similar? They could have had an adapter for the 0.00001% of people who wanted to use fibre. Also Intersil aren't a startup, they're older than Intel!
Ahh.. I remember the days of 1200 baud modems when they just came out. Cost $300 at the time, and one would wonder why anyone would need something faster. Why on earth would anyone even consider the dizzying speeds of a 2400 baud transfer?.
Nail
^
head
Uh ? Displays and TB have nothing to do with each other. 4K displays would be end user use case for DP 1.2. You know, that DisplayPort specification released in December 2009, more than 1 year ahead of Thunderbolt's first appearance ?
And has since 2009, same as DisplayPort. Support came with version 1.4 of the specification.
Thunderbolt brings nothing new to the table as far as 4K support is concerned. In fact, I doubt it can even do it, as current Thunderbolt uses a modified DisplayPort 1.1a specification, not the 1.2 specification.
EDIT : quick mention, DP 1.2 support, so support for 4K resolution, is coming with RedWood Ridge Thunderbolt controllers, slated to be released in 2013.
Does this mean existing Macs won't support 4K even though the DP specification dates back to 2009 ? Bummer.
They should implement also the great power-on feature (previously available on USB 1 and older Macs using USB 1 and even older ADB ones) to turn on Macs from the keyboard or dongles like the i-Cue. Great when the CPU is behind the desk or for disabled people:
USB Boot Dongle (i-Cue) for Mac
http://www.lindy.co.uk/usb-boot-dongle-i-cue-for-mac/32871.html
USB Boot Dongle (i-Cue) for Mac
http://www.lindy.co.uk/usb-boot-dongle-i-cue-for-mac/32871.html
It is actually the article itself that is steering the discussion in this direction by putting the two technologies right against each other in another misguided attempt to put a spin on the story ...
That said. The majority of people don't care about the underlying technology, they compare what they can do with a device or what marketing promised them they could do. In the case of Thunderbolt it was ultrafast external disks, external graphic card among others. Little to none has materialized so far .. thus the grudge.
T.
You guys realize that by 2014 thunderbolt will be probably sitting at 100Gbps right? A simple upgrade to thunderbolt will get it up to 20Gbps using current copper cables. The 100Gbps is a small update (chip) and simply using fiber optic cables which have already been announced by 2 vendors. If you're wondering why you need those speeds and have never run into that bottleneck, then it doesn't matter to you. For us who have run into the bottleneck, you know why.
Yeah, I think the current TB arrays are for this segment... but it's such a small, niche market that unless with broader adoption TB might be good as done.
Probably because HDMI is ubiquitous... it's a defacto standard for video output. TB on the other hand is pretty much a dead port ...
Cool, but what am I going to plug into the port?
There is a reason those Thunderbolt cables are $50 ...
They are a tuned transmission line with active termination at each end for data speed and reliability.
Try running data at that speed over a 3 meter cable and see what you end up with at the other end without that technology in the cable.
Exactly. We haven't seen 10Gbps USB3 cables yet. There is a good change they will require the same technology. That means they'll be pricey too.