Here's hoping we see some TB2 equipped laptops and a decent thunderbolt external GPU solution in 2014. Shame Intel is hell bent on preventing that though.
Love my job, since I've been bringing in $5600… I sit at home, music playing while I work in front of my new iMac that I got now that I'm making it online(Click on menu Home) http://goo.gl/cg8GjX
Dude $5600 a year. How do you survive on that? Man I was gonna click your link and give you my credit card # but after seeing the $5600, no thanks man.
Also I don't really want an Imac, you had said something better would have clicked your link MarcinRybus.
Hi Ian, I just want to point out that this is the second time you have misrepresented the flex I/O and USB configuration on a Gigabyte board preview. The NEC hubs are... hubs, not controllers. they therefore each need a connection to the PCH USB 3.0 ports. On the UD7 TH, they are using all 8 PCIe lanes, all 6 SATA 6G lanes, and all 4 remaining available USB 3.0 ports (2 of which go to NEC hubs for more physical USB 3.0 ports, and 2 of which get dedicated physical ports) for a total of 18 Flex I/O. You made a similar mistake in your preview of the Z87X-OC. I'll also point out that Gigabyte puts a comprehensive block diagram in the manual for every board. If you are ever confused about how things are routed I suggest checking the block digram before publishing. Thanks, g\
This isn't limited to Gigabyte boards. Many of his motherboard reviews are incorrect in this regard.
Look at the ASRock Z87 Extreme6/AC review. He calls the ASMedia hub a host controller, and it has the Flex I/O incorrect. It says 8/8/4 PCIe/SATA/USB3 whereas it should be 6/8/6 for that board.
Yeah, this is getting embarrassing now. The motherboard manufacturers have figured out that they can get away with USB 3.0 hubs because there are still so few USB 3.0 devices in circulation, but I would expect Anandtech to call them out on this.
And no ALC1150 on a board of this price? I guess corners had to be cut somewhere, but on the audio?
Apologies for not responding sooner, I have been at an overclocking meet all weekend and wrote this news on the train to the event.
I did not see your previous message (always easiest to contract me via email to make sure I do) but for whatever reason I had not made the connection between standard port hubs and hubs via a chip on board until your post. Describing 'hubs are hubs' threw me off a little, but basically it's similar to a PLX switch for USB. This ultimately limits upstream throughput, but there are very few instances where multiple USB devices are requiring different upstream and downstream transfers such that a hub should not be an issue (even more if they are simply IO devices). I will note it in future.
A note about the block diagram: it does not show the USB hubs as clearly as they should (chipset -> USB port -> USB hub -> 4x USB ports), but rather chipset -> USB hub -> 4 USB ports, which is unclear in that regard. Before I posted this I did request the chipset diagram from my PR, which did not show exactly what you said and left bits out to make it easier to understand. After returning to it after your post, I can see what you are saying and agree with you, however it is not 100% clear unless that controller/hub distinction is clear in ones mind.
As always, I am very much open to discussion if you feel I've misinterpreted anything - reviewing is always a two way process and I always aim to give our readers the best possible detail. This one was just a matter of hub/controller differentiation not being fixed in my understanding. Drop me a line anytime:).
Thanks Ian, I have very high standard for the site after years and years in your readership and so I don't like seeing errata. I agree that the diagram could be a bit clearer but that it would possibly seem a bit redundant to show a usb port branch before the hub as it is explicitly coming off the PCH and and explicitly not coming off the PCIe bus in the diagrams. So I think it is fairly clear. In any case keep up the good work and busy schedule, and if you find time feel free to correct the articles. I do agree that its a pretty smart move to use hubs onboard to increase USB port counts given the limitations of FlexIO and that for most uses the aggregate bandwidth issues of the hub won't be a problem. The I/O as such is better utilized on PCIe lanes and SATA ports. In the unlikely event I find corrections in the future I will email them to you instead of posting here. Thanks, g\
Motherboard manufacturers need to get their s**t together regarding USB 3.0 headers. Just bloody well group 'em together at the bottom of the board like the USB 2.0 headers! Move some of those fan headers and the useless COM header if you have to, just get this sorted out!
I'm not a big fan of USB 3 headers being on the bottom of the board. You can't use them when double width PCIe cards are in the bottom most slot. I prefer them to be on the right edge.
What's the point of Thunderbolt in Intel's eyes? Priced the way it is now it will never become a standard for anyone. They have to know that. USB is ubiquitous because you don't price yourself out of competition by acquiring the licensing. Faster connections are great, but it can't quadruple the price of your motherboard. 10-20 bucks extra, yeah, I could see that. But everything I've seen with Thunderbolt has been outrageously expensive. Damn shame.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
22 Comments
Back to Article
rhx123 - Saturday, December 14, 2013 - link
Here's hoping we see some TB2 equipped laptops and a decent thunderbolt external GPU solution in 2014.Shame Intel is hell bent on preventing that though.
Rishi100 - Saturday, December 14, 2013 - link
Still no HDMI 2.0?SunLord - Saturday, December 14, 2013 - link
HDMI 2.0 has only been released for 3 months it likely won't show up in devices till sometime in late 2014MarcinRybus - Saturday, December 14, 2013 - link
Love my job, since I've been bringing in $5600… I sit at home, music playing while I work in front of my new iMac that I got now that I'm making it online(Click on menu Home)http://goo.gl/cg8GjX
ThreeDee912 - Saturday, December 14, 2013 - link
Is there any kind of "Report Spam" button here? I've seen a couple of these comments still on some older posts too.alphasquadron - Sunday, December 15, 2013 - link
Dude $5600 a year. How do you survive on that? Man I was gonna click your link and give you my credit card # but after seeing the $5600, no thanks man.Also I don't really want an Imac, you had said something better would have clicked your link MarcinRybus.
melgross - Sunday, December 15, 2013 - link
Nah, they're usually talking about weekly, or monthly income. Still a scam.Androidtech - Wednesday, December 18, 2013 - link
That is pretty sad I make more than that playing my PS4 on my new Samsung Oled TV just off the interest in my savings accounts every month.genzai - Saturday, December 14, 2013 - link
Hi Ian,I just want to point out that this is the second time you have misrepresented the flex I/O and USB configuration on a Gigabyte board preview. The NEC hubs are... hubs, not controllers. they therefore each need a connection to the PCH USB 3.0 ports. On the UD7 TH, they are using all 8 PCIe lanes, all 6 SATA 6G lanes, and all 4 remaining available USB 3.0 ports (2 of which go to NEC hubs for more physical USB 3.0 ports, and 2 of which get dedicated physical ports) for a total of 18 Flex I/O. You made a similar mistake in your preview of the Z87X-OC. I'll also point out that Gigabyte puts a comprehensive block diagram in the manual for every board. If you are ever confused about how things are routed I suggest checking the block digram before publishing.
Thanks,
g\
The Von Matrices - Saturday, December 14, 2013 - link
This isn't limited to Gigabyte boards. Many of his motherboard reviews are incorrect in this regard.Look at the ASRock Z87 Extreme6/AC review. He calls the ASMedia hub a host controller, and it has the Flex I/O incorrect. It says 8/8/4 PCIe/SATA/USB3 whereas it should be 6/8/6 for that board.
The Von Matrices - Saturday, December 14, 2013 - link
Whoops, I meant he said 6/8/4 PCIe/SATA/USB3 in the review whereas it actually should be 6/6/6.The_Assimilator - Saturday, December 14, 2013 - link
Yeah, this is getting embarrassing now. The motherboard manufacturers have figured out that they can get away with USB 3.0 hubs because there are still so few USB 3.0 devices in circulation, but I would expect Anandtech to call them out on this.And no ALC1150 on a board of this price? I guess corners had to be cut somewhere, but on the audio?
The Von Matrices - Saturday, December 14, 2013 - link
Part of the advantage of USB 3 hubs on Z87 is that it solves the C1 chipset errata of USB devices disappearing upon resume from sleep.IanCutress - Monday, December 16, 2013 - link
Apologies for not responding sooner, I have been at an overclocking meet all weekend and wrote this news on the train to the event.I did not see your previous message (always easiest to contract me via email to make sure I do) but for whatever reason I had not made the connection between standard port hubs and hubs via a chip on board until your post. Describing 'hubs are hubs' threw me off a little, but basically it's similar to a PLX switch for USB. This ultimately limits upstream throughput, but there are very few instances where multiple USB devices are requiring different upstream and downstream transfers such that a hub should not be an issue (even more if they are simply IO devices). I will note it in future.
A note about the block diagram: it does not show the USB hubs as clearly as they should (chipset -> USB port -> USB hub -> 4x USB ports), but rather chipset -> USB hub -> 4 USB ports, which is unclear in that regard. Before I posted this I did request the chipset diagram from my PR, which did not show exactly what you said and left bits out to make it easier to understand. After returning to it after your post, I can see what you are saying and agree with you, however it is not 100% clear unless that controller/hub distinction is clear in ones mind.
As always, I am very much open to discussion if you feel I've misinterpreted anything - reviewing is always a two way process and I always aim to give our readers the best possible detail. This one was just a matter of hub/controller differentiation not being fixed in my understanding. Drop me a line anytime:).
Ian
genzai - Monday, December 16, 2013 - link
Thanks Ian,I have very high standard for the site after years and years in your readership and so I don't like seeing errata. I agree that the diagram could be a bit clearer but that it would possibly seem a bit redundant to show a usb port branch before the hub as it is explicitly coming off the PCH and and explicitly not coming off the PCIe bus in the diagrams. So I think it is fairly clear. In any case keep up the good work and busy schedule, and if you find time feel free to correct the articles. I do agree that its a pretty smart move to use hubs onboard to increase USB port counts given the limitations of FlexIO and that for most uses the aggregate bandwidth issues of the hub won't be a problem. The I/O as such is better utilized on PCIe lanes and SATA ports. In the unlikely event I find corrections in the future I will email them to you instead of posting here.
Thanks,
g\
MadAd - Saturday, December 14, 2013 - link
$430!!!!!!!!!The_Assimilator - Saturday, December 14, 2013 - link
Motherboard manufacturers need to get their s**t together regarding USB 3.0 headers. Just bloody well group 'em together at the bottom of the board like the USB 2.0 headers! Move some of those fan headers and the useless COM header if you have to, just get this sorted out!The Von Matrices - Saturday, December 14, 2013 - link
I'm not a big fan of USB 3 headers being on the bottom of the board. You can't use them when double width PCIe cards are in the bottom most slot. I prefer them to be on the right edge.skyntax.256 - Saturday, December 14, 2013 - link
this motherboard is Cerberus colored. Very cool.DV8_MKD - Monday, December 16, 2013 - link
All these comments and not one about TB 2. Talk about the new firewire...Hrel - Monday, December 16, 2013 - link
What's the point of Thunderbolt in Intel's eyes? Priced the way it is now it will never become a standard for anyone. They have to know that. USB is ubiquitous because you don't price yourself out of competition by acquiring the licensing. Faster connections are great, but it can't quadruple the price of your motherboard. 10-20 bucks extra, yeah, I could see that. But everything I've seen with Thunderbolt has been outrageously expensive. Damn shame.