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.
Thunderbolt Adoption Reportedly Slowed by Intel's Licensing and Certification, But Improvements Coming
- Thread starter MacRumors
- Start date
- Sort by reaction score
Awesome! Why would I use my 15" MBP's dedicated Radeon video card and connect it to my 30" monitor via DisplayPort, when I can use a $64 USB video card. Clearly they should get rid of Thunderbolt altogether since it's so useless.
Sarcasm aside, I know there are USB video cards. But that's not how you want to connect your main external display in most cases.
The funny thing is that for many years FW800 offered twice as much speed compared to USB 2 but despite what you describe as a negligible price gap between the two, 95% of the population turned out to be cheapskates that did not want to pay that negligible price difference (and no, it wasn't about interoperability, very few people share HDDs with other people and as you said almost all FW devices had a USB port as well).
Think about that, 95% of the population is not willing to pay $30 extra to get twice the speed.
----------
I was being sarcastic. And I obviously meant PCI-e slots since we talk here about ports to plug stuff in. And even if maybe most PC desktops had PCI-e slots, what counts in the TB doomsdayers mind is people actually buying peripherals that plug into a port.
I don't know what world you were living in, but when I bought my portable computer powered 500GB backup drive for my MBP in 2008 with FW800 and USB 2.0, it cost more like $100 more than a USB 2.0 only drive ($250 vs $150 if I recall correctly or something close to that. It sure as heck wasn't wasn't $30 more or when I bought a second drive for my Netbook I wouldn't have got the USB 2.0 only drive despite the fact the Netbook didn't have Firewire since I could potentially re-use it at some point in the future on one of my other Macs, if needed).
FW400 drives were often maybe $40-60 more 4-6 years ago, but FW800 was always a good premium over both. In short, they weren't giving it away. It had a limited market and they knew you were getting a big speed improvement to boot. Perhaps NOW you might find some FW800 drives closer to $30 more in SOME cases compared to certain drives, but I sure as heck never saw any at that price differential when I was shopping for one in 2008.
I found this statement very strange. I have used an Asus monitor (which I returned), and currently have connected both my Dell & Samsung monitors at home to my '12 Mac Mini with no issues. That's already three major brands there. It doesn't seem like HDMI should be an issue with OSX.
It shouldnt be, but it is. It's well documented too.
HDMI on the rMBP is just dodgy. If I plug one of my 2 monitors into it, it picks them up as a TV with a 720p resolution. If I plug them in via a DVI > Displayport cable, they work fine.
It looks like its not just the rMBP either. The Macbook Air and Mini have the same problem. Someone at Apple messed up.
It's not just me:
http://www.ehmac.ca/mac-iphone-ipad...ting/100898-rmbp-external-screen-quality.html
https://forums.macrumors.com/threads/1488630/
https://forums.macrumors.com/threads/1460607/
https://forums.macrumors.com/threads/1390034/
https://discussions.apple.com/thread/4249436
https://discussions.apple.com/thread/4119545
https://discussions.apple.com/thread/4080525
It's a really hard error/issue to describe. The only way to describe it would be to get you to plug your mac into your TV set. You'll notice that there is a noticeable difference in quality, especially with text. Now imagine that lack of detail magnified (it seems to get worse the smaller the screen), to the point where text is barely readable. Thats what it looks like.
Hence why I feel it is an utter Failure...USB3 however is making leaps and bounds.
TB is too expensive, very limited, only works on Macs (even though that is all I have is Macs) and I just think it is not worth the cost, especially with USB 3.2 going to 10GB by 2014.
Very few vendors for TB also, seems LaCie has the hold on the market for some reason? I use Seagate and they just came out with adapters for it, so prefer to get one of those for $99 and use my current HDD then buy a $300 500GB LaCie HDD
'Pro-sumer" what cr-p! Then why is it on my Macbook Air? This technology is dead! Apple blew it. Just look at the price on Thunderbolt peripherals, Adapters, and hard drives.
I am a con-sumer.
Apple blowing it and putting it on consumer focused models does not mean the technology is dead.
Intel has been pushing TB has a prosumer level technology since 2011, I linked to their stance then multiple times already. That Apple tried to push it down to consumers and failed (in light of it being more prosumer than consumer friendly) is not a failure of the tech, nor does it mean the tech is dead in the segment its meant for.
Thunderbolt is doing fine in the prosumer sector.
I just checked prices for two examples:
- 2.5" cases, random vendor (ICY BOX), USB 2-only: $32, FW800: $65
- Voyager dock: USB 3-only: $38, FW800: $78
In particular in the case of the latter, ie, a dock, you pay $40 once (or twice if you need two docks) which is fairly little in the grand scheme of things of owning a computer.
I guess you missed the part where I said I was talking about OVER FOUR YEARS AGO when it was actually a relevant format being one of the faster ones available. In fact, I mentioned this several times and said you might find one now for that difference, but certainly not back then and certainly not pre-packaged. I mean who cares what they are going for NOW when it's now outdated and slow as heck compared to USB 3.0 (and yet STILL costs more!) What do you want, 800Mbits for $78 or 5Gbits for $38? I'll take the latter, thank you. There isn't a single TB enclosure available that I can find to compare there, but I'm sure it'd be more like $178 if it were available.
As per usual
"technically better" doesn't always win
These idiots still don't get it
VHS vs BETAMAX
Sure, beta was "better" but how'd that go?
Total arrogance and lack of respect of history
----------
So true- apple fan-frauds always fighting reality.......
"technically better" doesn't always win
These idiots still don't get it
VHS vs BETAMAX
Sure, beta was "better" but how'd that go?
Total arrogance and lack of respect of history
----------
So true- apple fan-frauds always fighting reality.......
Question: is Thunderbolt I/O support built into Windows 8? Does Intel offer software drivers so it works under Windows 7? If the answer to both is yes, then Thunderbolt adoption could roll out fairly quickly, especially now that we're about to get optical Thunderbolt connections (which would finally put the e-SATA standard to rest).
Thunderbolt and USB are complementary technologies. USB will never become a native digital display interface, whereas Thunderbolt ports use the VESA mini-DisplayPort connector and can operate in DisplayPort legacy signaling mode. USB 3.0 is backwards compatible with the billions of USB devices in the world, and Thunderbolt can drive just about any display that exists. Thunderbolt was meant as a replacement for ExpressCard slots, VGA, DVI, HDMI, DisplayPort and FireWire ports on Ultrabooks, notebooks, all-in-one and USFF PCs, not as a replacement for USB.
This is not a format war. AMD and Intel have both integrated USB 3.0 controllers into their chipsets, which means virtually every x86 PC produced from here on out will have USB 3.0. Having Thunderbolt is a bonus. Apple largely buried the cost of the controllers so that Mac prices did not increase when they introduced this technology. Having that extra functionality available is not a bad thing. For those who keep saying USB 3.0 is going to "win", well then I guess owners of Macs with both USB 3.0 and Thunderbolt ports are "bi-winning".
Most people do not need to buy a semi-tractor trailer or diesel locomotive because an automobile serves their purposes much better and costs considerably less. However, those same people could not enjoy their current standard of living if those other modes of transportation ceased to exist. Most PC end users do not require their devices to include 40 or 100 GbE capabilities, but they depend on a larger network that uses those technologies extensively. Most folks may not yet have a need for a 10 Gbit/s I/O interface on their PC either, but they do like watching movies, listening to music and the fruits of various other industries that are seeing widespread adoption of Thunderbolt devices. Popularity is not the only benchmark of success.
Apple is the largest purchaser of semiconductors in the world, and it is reasonable to believe they are Intel's number one customer in the consumer space. Due to the cost of the technology, Light Peak was really only sought after by two OEMs, Apple and Sony, and only Apple was willing to go all-in. Apple is therefore the only significant customer of Thunderbolt controllers. When discussing Thunderbolt, one needs to bear in mind that it is a proprietary interface developed by Intel for Apple. As long as Apple chooses to keep buying controllers from Intel, the interface will continue to exist. Saying Thunderbolt is DOA demonstrates a general lack of understanding of the underlying economics. For the past several years, Mac sales have grown steadily while the rest of the PC market has stalled and actually started to contract. Why would Apple eliminate a feature that differentiates them from the competition when those features clearly seem to be allowing Apple to outperform the industry?
Based on its per channel data rate, Thunderbolt is the fastest interface to ever appear in a consumer device. It is a sort of Hayabusa moment for PC I/O. It is also the least expensive 10 Gbit/s per channel solution in terms of cost per port by a factor of 5 or more. It may be expensive, but it is not necessarily overpriced. While increased adoption will introduce some economies of scale, and moving to newer processes along with better integration of some of the functions at the edges will help to lower prices, Thunderbolt will always be a premium technology; this fundamentally can't be changed. The controller is huge and costs almost as much as the PCH on most systems, and the transceivers are operating close to the maximum speeds currently achievable even in the enterprise space. The reason why Ferrari's aren't $14k is not because they can't sell enough of them to hit that price point.
The problem is that you are not driving your display at its native resolution, with 1:1 pixel mapping. This is not necessarily the fault of the Mac's HDMI port at all. HDMI was designed for HDTVs, not high resolution PC displays. Unless your display has an HDMI 1.3 or newer sink device in it, it can only support resolutions up to 1920x1200 at 60 Hz. It will therefore report that as the maximum supported resolution to whatever is driving it. Displays that can support resolutions higher than 1920x1200 @ 60 Hz over HDMI will still require a "High Speed" HDMI cable. Both the Mac and the display will also have overscan/underscan settings that need to be configured to get 1:1 output. The Mac could be outputting a signal that is spot on, but the scaler built in to the display is mangling it. There are also various ways to force output of a resolution that is not expressly listed in your display's EDID, if it comes down to that.
This is why DisplayPort is preferable to HDMI for connecting displays to a PC.
I've been buying FW800 gear since '04, and while I've paid dearly for some items, others were quite reasonable. I have several USB 2.0/FW400/FW800 enclosures that I bought for $30-$40 apiece. I also picked up a 2TB WD USB 2.0/FW400/eSATA drive at Best Buy when all I needed was a bare drive because they were the same price. (I know, not a FW800 example, and it was early 2009, but it was the good ol' days before the flood in Thailand.) I think part of the problem with FW hard drive pricing is that it was perceived solely as a Mac feature, and certain brands like LaCie and G-Technology became very prominent while putting out high-margin, form over function designs. This obviously continues today and has spread to Thunderbolt drives. In the end, it's the cost of the controllers that determines the lower end of the pricing scale. The Oxford Semiconductor controllers that were the most popular choice for FW800 gear need only have added $25 or so to the retail price of a drive as compared to just USB 2.0.
Thunderbolt per se doesn't so much require drivers. The connected devices appear simply as PCIe devices to the host. In order for a product to get Thunderbolt certification for Windows, however, the manufacturer has to provide drivers that can support hot-plugging for the PCIe devices used in that product.
I think USB 3.0 with UASP has already put eSATA to rest for the most part. And although optical cables for Thunderbolt will probably become more widely available than they are today, I don't see Thunderbolt switching to optical connections until 2017 or so.
No cable will make existing Thunderbolt controllers faster. Intel's roadmap appears to call for a speed increase to 20 Gbit/s per channel in 2014. Copper cables and connectors will likely continue to be the norm on the desktop and should support the new speed just fine. Building optical transceivers into devices when they aren't needed would be a waste of money, circuit board real-estate and power budget for everyone except the very small minority that require a link longer than 3.0 m. Keeping the transceivers in the cables allows for multiple transmission media to be used and reduces the overall cost for everyone.
This is not a format war. AMD and Intel have both integrated USB 3.0 controllers into their chipsets, which means virtually every x86 PC produced from here on out will have USB 3.0. Having Thunderbolt is a bonus. Apple largely buried the cost of the controllers so that Mac prices did not increase when they introduced this technology. Having that extra functionality available is not a bad thing. For those who keep saying USB 3.0 is going to "win", well then I guess owners of Macs with both USB 3.0 and Thunderbolt ports are "bi-winning".
Most people do not need to buy a semi-tractor trailer or diesel locomotive because an automobile serves their purposes much better and costs considerably less. However, those same people could not enjoy their current standard of living if those other modes of transportation ceased to exist. Most PC end users do not require their devices to include 40 or 100 GbE capabilities, but they depend on a larger network that uses those technologies extensively. Most folks may not yet have a need for a 10 Gbit/s I/O interface on their PC either, but they do like watching movies, listening to music and the fruits of various other industries that are seeing widespread adoption of Thunderbolt devices. Popularity is not the only benchmark of success.
Apple is the largest purchaser of semiconductors in the world, and it is reasonable to believe they are Intel's number one customer in the consumer space. Due to the cost of the technology, Light Peak was really only sought after by two OEMs, Apple and Sony, and only Apple was willing to go all-in. Apple is therefore the only significant customer of Thunderbolt controllers. When discussing Thunderbolt, one needs to bear in mind that it is a proprietary interface developed by Intel for Apple. As long as Apple chooses to keep buying controllers from Intel, the interface will continue to exist. Saying Thunderbolt is DOA demonstrates a general lack of understanding of the underlying economics. For the past several years, Mac sales have grown steadily while the rest of the PC market has stalled and actually started to contract. Why would Apple eliminate a feature that differentiates them from the competition when those features clearly seem to be allowing Apple to outperform the industry?
Based on its per channel data rate, Thunderbolt is the fastest interface to ever appear in a consumer device. It is a sort of Hayabusa moment for PC I/O. It is also the least expensive 10 Gbit/s per channel solution in terms of cost per port by a factor of 5 or more. It may be expensive, but it is not necessarily overpriced. While increased adoption will introduce some economies of scale, and moving to newer processes along with better integration of some of the functions at the edges will help to lower prices, Thunderbolt will always be a premium technology; this fundamentally can't be changed. The controller is huge and costs almost as much as the PCH on most systems, and the transceivers are operating close to the maximum speeds currently achievable even in the enterprise space. The reason why Ferrari's aren't $14k is not because they can't sell enough of them to hit that price point.
The problem is that you are not driving your display at its native resolution, with 1:1 pixel mapping. This is not necessarily the fault of the Mac's HDMI port at all. HDMI was designed for HDTVs, not high resolution PC displays. Unless your display has an HDMI 1.3 or newer sink device in it, it can only support resolutions up to 1920x1200 at 60 Hz. It will therefore report that as the maximum supported resolution to whatever is driving it. Displays that can support resolutions higher than 1920x1200 @ 60 Hz over HDMI will still require a "High Speed" HDMI cable. Both the Mac and the display will also have overscan/underscan settings that need to be configured to get 1:1 output. The Mac could be outputting a signal that is spot on, but the scaler built in to the display is mangling it. There are also various ways to force output of a resolution that is not expressly listed in your display's EDID, if it comes down to that.
This is why DisplayPort is preferable to HDMI for connecting displays to a PC.
I've been buying FW800 gear since '04, and while I've paid dearly for some items, others were quite reasonable. I have several USB 2.0/FW400/FW800 enclosures that I bought for $30-$40 apiece. I also picked up a 2TB WD USB 2.0/FW400/eSATA drive at Best Buy when all I needed was a bare drive because they were the same price. (I know, not a FW800 example, and it was early 2009, but it was the good ol' days before the flood in Thailand.) I think part of the problem with FW hard drive pricing is that it was perceived solely as a Mac feature, and certain brands like LaCie and G-Technology became very prominent while putting out high-margin, form over function designs. This obviously continues today and has spread to Thunderbolt drives. In the end, it's the cost of the controllers that determines the lower end of the pricing scale. The Oxford Semiconductor controllers that were the most popular choice for FW800 gear need only have added $25 or so to the retail price of a drive as compared to just USB 2.0.
Thunderbolt per se doesn't so much require drivers. The connected devices appear simply as PCIe devices to the host. In order for a product to get Thunderbolt certification for Windows, however, the manufacturer has to provide drivers that can support hot-plugging for the PCIe devices used in that product.
I think USB 3.0 with UASP has already put eSATA to rest for the most part. And although optical cables for Thunderbolt will probably become more widely available than they are today, I don't see Thunderbolt switching to optical connections until 2017 or so.
No cable will make existing Thunderbolt controllers faster. Intel's roadmap appears to call for a speed increase to 20 Gbit/s per channel in 2014. Copper cables and connectors will likely continue to be the norm on the desktop and should support the new speed just fine. Building optical transceivers into devices when they aren't needed would be a waste of money, circuit board real-estate and power budget for everyone except the very small minority that require a link longer than 3.0 m. Keeping the transceivers in the cables allows for multiple transmission media to be used and reduces the overall cost for everyone.
Last edited:
Blowing it is a strong description. They overly marketed it when as far I can see their primary intention was to release a display + dock combination and a small amount of external storage based on it. The rMBP models and imac may have 2 ports. The others just have it where mini displayport already existed. It builds out on an existing port. I'm not inclined to believe that if it wasn't added, we would have received something else in its place.
I get what you're saying, and do agree with most parts. I dont pretend to be an expert in how HDMI works but here's what I dont get. If I drop into bootcamp running Windows or Linux (Debian) - it works fine. It's only in OS X that it cant figure out the resolution.
It's not a major issue for me as I'm happy to use the two thunderbolt ports, it's just a bit frustrating that you've paid £1500 for a laptop and one of its 'features' doesn't work as expected.
How would optical TB connections help adoption at all ? The cabling is going to be massively more expensive than the current, already too expensive options and it won't be faster.
Why would optical TB "finally" put eSATA to test ? eSATA is 6 Gbps vs TB's 10 Gbps...
I don't understand how you're thinking here.
Unfortunately, it's not just a case of soldering the Thunderbolt wires onto a bog standard PCIe backplane...
Thing is, for some professionals, the ability to plug a Fibre Channel or pro capture card into their laptop is going to be transformative and worth the $$$howmuch!? price tag.
However, the typical consumer simply doesn't need PCIe these days. Back in the day, when I assembled my own PCs, the PCI/ISA bus would be loaded up with sound card, ethernet card, maybe a second disc controller, SCSI card for a scanner, Firewire card for video editing etc. However, increasingly that all ended up on the motherboard or available via USB and PCI sat there unused. The only thing I've plugged into a PCI slot in the last 5-6 years is a TV tuner (and some of those turn out to be the maker's USB product stuck onto a USB controller card!)
So PCIe enclosures are always going to be a small market, and in electronics, small market == high prices.
...except its not a port replicator or hub. It is a brand new set of controllers over and above what the laptop provides.
The big cock-up there was they've taken too long to get to market - if the Belkin hub had been available last summer, with eSATA and USB3 as then promised, it would have been a perfect upgrade for 2011 Macs. Limping out next September, sans eSATA, when Macs with USB3 are becoming commonplace and those 2011 Macs are ripe for replacement... not so much.
I'm not so sure. Thunderbolt is pretty irrelevant to the generic desktop PC (since they have internal PCIe slots) or big blocky laptops with loads of ports and/or ExpressCard slots. It is mainly a tech for high-end-but-slim laptops and small-form-factor/all-in-one computers - Apple have a huge chunk of that particular market segment, and a deep-pocketed customer base. If that market isn't big enough then Thunderbolt has problems.
Thats why I think Apple should start making more Thunderbolt peripherals itself - starting with the proverbial decent Thunderbolt dock - in nice Jonny Ive cases, and priced to encourage third parties to undercut them... I'm sure its easier to predict how many units you would sell if Apple have established a market and your product is 20% cheaper...
Or you could just unmount the HDD when not using it. If the data needs to be accessed from other computers on your net work just share the drive?
So you're saying people have to work around daisy chaining... I really don't understand what people feel is so great about daisy chaining devices. I thought we moved away from such bus topologies onto star topologies because it's easier to add and remove devices dynamically without impacting the rest of the connected devices.
10Base2 Ethernet anyone ? SCSI chains ? You really would rather go back to this stuff than what replaced it ?
yes actually because I find it more convenient. TB offers me a on cable solution for a portable. One TB cable connects all my external storage and displays.
One cable ? Sorry, that's not sufficient. Every device in the chain still requires its own cable.
Daisy chaining died for good reasons in the 90s. Lets leave it buried. Anyway, Thunderbolt is also capable of being used in a star topologie.
And with the exception of the rather expensive direct PCIe drives, the drive is on a 6 Gbps or slower SATA link anyway....
It's capable, but it cannot be used in a star topology.
