The Nokia n1 tablet might just be the first device shipping (it is shipping) with Type-C. On the chip side, as far as i know Cypress was the first to have one.
I expect that to be the case for most phones/tablets. The USB C connector support USB 3 and 3.1 speeds but is actually a separate thing and doesn't require USB3 support. It can be used simply as a new connector for usb2 to replace Micro USB with something that is reversible. USB3 support has been pretty limited on phones and tables and the few Samsung phones I've read about that had USB3 also had issues with it that caused them to default to having it off and requiring you to manually enable it when you wanted to use it.
type-c limited to usb2 speeds still makes sense though, since you don't really need the full 10gbit and 100w usb3.1 tops out at. 2a/10w would be enough to charge it, transfer rates would be capped at 30mbyte/s though, but it's not that you would get that much more with usb3.1 anyways.
Regarding the power issue, seems like they should be able to add a PCIE power connection into the back-panel to provide the power, rather than trying to wire the extra 100W through the motherboard PCB. Most builds have extra PCIE power connectors that can be utilized to provide the extra power.
The 100W option requires extra hardware on the Mobo. It's 5A @ 20V; meaning that an extra set of power circuitry is needed to create a 20V source on the mobo to output it.
...and then a second one to somewhere along the front/bottom edge for headers connecting front panel hookups. No, if it's going to happen it needs to be integrated into the mobo somehow.
And while personally, I lean towards hoping the 20V bit dies along with the rest of the USB groups one cable to bind them fantasies. (If you've got dozens of mutually incompatible not-USB streams being sent over the same type of cable, it's no longer any more universal than RS-232 was 25 years ago.)
If it does happen, my hope is that it finally prods the ATX people into replacing the 24 pin legacy connector with something that ditches the huge number of 3.3/5V wires that a 90's era PC (where everything ran on 3.3 or 5V) needed with a somewhat smaller connector that in addition to giving a bit of 20V has enough 12v that low power systems could dispense with the extra 4pin connector as well.
For gaming/workstation class systems, an extra 12V connector probably is inevitable unless we make the new main ATX connector as huge as the current one is.
The current connector has 23 pins in use (a 24th, providing -5V for ISA cards was removed years ago). In terms of removing pins the best we can do is probably to remove all 4 3.3v pins (legacy PCI), the -12v pin (RS232), 4/5 5v pins (need 1 for low power USB) and 5/8 ground pins (one each for the 2 12v pins and 1 remaining 5v pin); this reduces the pinout down to 9.
Going back up to add support for more modern uses, a single 20V pin, 2 12v pins, and 2 more ground pins (5 shared among 4x12v, 1x5v, 1x20v); brings us back up to 14 pins. 14's already getting fat enough that it's going to be getting annoying to bend. Adding 4 more pins replace the other half of the 8 pin connector would bump it back to 18; almost as bad as the original 20pin cable and mostly defeating the purpose. With 8 + 4 pin (and occasionally 8 + 8 or * + 4 + 6 (pcie)) additional 12V connectors becoming common on high end mobos; we'd still be connecting multiple cables at the top end.
Going the other direction, only adding a 20v wire to the reduced cable and stopping at 10 pins would make the cable easier to route; but would continue to impose a second power cable on low end mITX boards - the 4pin is ubiquitous even on boards with low power BGA processors and only pcie 1x (25W) slots - where cable routing is most difficult.
I suggested adding 2 12v wires to the redesigned main cable as a reasonable in compromise position. It still gives the big cable a badly needed diet to make routing easier. Low end systems are able to eliminate the 2nd cable entirely. And on the top end, it will at least serve as a counterbalance to the continuing drive for yet another 12v cable; and maybe take a step back on mainstream gaming boxes.
I'm disappointed that it is a single USB c connector as it wastes so much space. In addition to being easier to connect you should be able to fit more ports on the back panel so you don't have 20 usb headers you need to connect to use all the usb ports the board supports.
Why are you disappointed? It's too early to replace everything with USB 3.1 right now, there's not enough industry support to do this. The chipsets are not ready, issues with the power delivery, connector designers and so on.
It's going to take a few years before USB 3.1 have the proper support from everybody.
MSI is just trying to be the first with USB 3.1 connector and they did it by doing this weird setup with third party chipset, combining ports and so on.
USB C connector does not mean USB 3.1. It is entirely possible to do USB C ports that are USB 3.0 or even USB 2 and there are plenty of USB 2 and 3 ports supported by the chipset for them to not all fit on the back panel.
as often with first gen products, oems just wanna "get it out there" first, so one lonely port is enough to slap that "usb3.1" sticker on the box. more ports wouldn't make much sense anyways, if the bandwidth to the chipset is limited. i just can't wait until intel and amd get off their asses and give us glorious usb3.1 integrated into the chipset per default.
I just wish that they would abandon their love affair with Qualcomm NICs. I'd rather see them feature something like Intel server grade NICs on their motherboards - that would be a real upgrade.
Otherwise, I'm pretty happy with MSI.
I actually have become fond of their XPower series of motherboards. Their big weak point seems to be RAM overclocking (fewer profiles) and their UI isn't always the best. Oh, and Command Center has been buggy at times.
Their Lightning GPUs are among the best. I just hope they release big Maxwell and whatever AMD offers with top notch VRAM (or whatever HBM equal it is) this time (they had to use Elpida on the 780L, although they later replaced with Hynix).
Otherwise pretty good. Launch day BIOS stability is not good, but they do seem good about releasing BIOS updates and they have good tech support.
Intel currently controls the PCH, and with that, there's limited room for motherboard innovation, but I am hoping to see something interesting in the upcoming years, especially now that single threaded performance doesn't seem to be getting much faster with each generation.
Come to think of it.. I could have misread your comment.. I was talking about the 3.0 header on the board that you connect your USB 3.0 case cable to..
i guess it's just cheaper to implement a simple micro usb2.0 port right now and since that's good enough for charging and the occasional data transfer, it will take a while until type-c is ubiquitous on mobile devices.
another question is how this will be reglemented in the EU, since right now phones have to come with micro usb and i don't know if you're even allowed to sell a phone with type-c here at the moment.
I believe that Apple supply a Lightning-to-microUSB adapter in European retail boxes so I presume that any manufacturer who wishes to implement USB 3.1 Type-C would do similarly.
Also, Ian Cutress, Dong Ngo has reported that even the X99A Gaming 9 ACK motherboard (and GT72 gaming notebook) will come with a Type-C port:
I've liked the look of MSI's motherboards. I have a H55 socket 1156 still running for me with a Core i3.
I have two of their Twin Frozr R9 290 cards. I have been pretty happy with them. The fans on them are not 24/7 high rpm and will freeze up after spurting a little oil from the fan internals. I've read online that they can have mineral oil added to them and work as new at that point.
I handled this through warranty. It cost me $18 to ship. MSI's policy says it can take up to 28 days in RMA, but I got mine back in a week, but it wasn't the same card. The card wasn't as new as mine and the part that screws to the case was bent. I had to bend it back to get it to mount into the board and case. I had RMAed a Gigabyte GTX660 and was without the card for about 17 days.
With the exception of laptop/desktop computers (where the data on the drive matters); you almost never get the same item you sent in for RMA back. To minimize the turnaround they send you a replacement off their shelves immediately and then shelve yours after fixing it (or determining it's irreparable and disposing of it).
I'm mildly disappointed that they made you send the whole card in for a round trip though. A few years back I had a fan die on my XFX(?) 5870. I filed an RMA stating that the fan refused to spin and the card was overheating. They asked if I felt comfortable unscrewing the existing heat sink and replacing it myself. When I said yes, they just mailed me a replacement and didn't require me to return the original.
How is it difficult to get 100 watts to each and every USB motherboard port? Currently pcie x16 slots are required to supply 75 watts each. Many motherboards now have 3-6 of these slots. If you need more power you just plug in another power connector. The biggest difficulty I can see is OEM systems using within 50 watts of the max power output of the power supply. Only in that event would I see it being a difficulty. If it's too difficult to get that much power to USB on the motherboard backplate designers could simply have USB headers that are cabled to a bracket that would fit in the back of a computer and a place on the inside of the bracket to plug in a few power connectors. I predict this will lead to a need for a way for a power supply to communicate with the computer itself to state how much power it has available.
Why do you need every port toi be 100W capable? Realistically only 1 or 2 ports will need such power, and if the motherboard has to provide 100W interface to say 12USB 3.0 ports in the future, the cost involved will be prohibitory. And normal power supplies will not be able to handle even 6 USB3.0 ports each drawing 100W. So what is the point of wiring up all USB ports for power when 1) you only need a few 2) power supply limits?
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.
32 Comments
Back to Article
jjj - Monday, January 12, 2015 - link
The Nokia n1 tablet might just be the first device shipping (it is shipping) with Type-C.On the chip side, as far as i know Cypress was the first to have one.
jjj - Monday, January 12, 2015 - link
oups, it appears that the Nokia is using the reversible connector but on USB2 lol.kpb321 - Monday, January 12, 2015 - link
I expect that to be the case for most phones/tablets. The USB C connector support USB 3 and 3.1 speeds but is actually a separate thing and doesn't require USB3 support. It can be used simply as a new connector for usb2 to replace Micro USB with something that is reversible. USB3 support has been pretty limited on phones and tables and the few Samsung phones I've read about that had USB3 also had issues with it that caused them to default to having it off and requiring you to manually enable it when you wanted to use it.fokka - Tuesday, January 13, 2015 - link
type-c limited to usb2 speeds still makes sense though, since you don't really need the full 10gbit and 100w usb3.1 tops out at. 2a/10w would be enough to charge it, transfer rates would be capped at 30mbyte/s though, but it's not that you would get that much more with usb3.1 anyways.BubbaJoe TBoneMalone - Monday, January 12, 2015 - link
Whatever happened to the ASUS USB 3.1 motherboards codenamed Shadowcat and Quicksilver from Computex 2014?http://rog.asus.com/325712014/asus-gaming-motherbo...
7amood - Tuesday, January 13, 2015 - link
There is no type-C on thoseJBVertexx - Monday, January 12, 2015 - link
Regarding the power issue, seems like they should be able to add a PCIE power connection into the back-panel to provide the power, rather than trying to wire the extra 100W through the motherboard PCB. Most builds have extra PCIE power connectors that can be utilized to provide the extra power.DanNeely - Monday, January 12, 2015 - link
The 100W option requires extra hardware on the Mobo. It's 5A @ 20V; meaning that an extra set of power circuitry is needed to create a 20V source on the mobo to output it.Gigaplex - Monday, January 12, 2015 - link
You can do that on a little daughterboard on the back panel, it doesn't have to route through the main part of the motherboard.DanNeely - Tuesday, January 13, 2015 - link
...and then a second one to somewhere along the front/bottom edge for headers connecting front panel hookups. No, if it's going to happen it needs to be integrated into the mobo somehow.And while personally, I lean towards hoping the 20V bit dies along with the rest of the USB groups one cable to bind them fantasies. (If you've got dozens of mutually incompatible not-USB streams being sent over the same type of cable, it's no longer any more universal than RS-232 was 25 years ago.)
If it does happen, my hope is that it finally prods the ATX people into replacing the 24 pin legacy connector with something that ditches the huge number of 3.3/5V wires that a 90's era PC (where everything ran on 3.3 or 5V) needed with a somewhat smaller connector that in addition to giving a bit of 20V has enough 12v that low power systems could dispense with the extra 4pin connector as well.
fokka - Tuesday, January 13, 2015 - link
oh please yes. the 24pin connector is a monstrum, plus needing an additional 4/8pin for the cpu doesn't make it any better neither.DanNeely - Wednesday, January 14, 2015 - link
For gaming/workstation class systems, an extra 12V connector probably is inevitable unless we make the new main ATX connector as huge as the current one is.The current connector has 23 pins in use (a 24th, providing -5V for ISA cards was removed years ago). In terms of removing pins the best we can do is probably to remove all 4 3.3v pins (legacy PCI), the -12v pin (RS232), 4/5 5v pins (need 1 for low power USB) and 5/8 ground pins (one each for the 2 12v pins and 1 remaining 5v pin); this reduces the pinout down to 9.
Going back up to add support for more modern uses, a single 20V pin, 2 12v pins, and 2 more ground pins (5 shared among 4x12v, 1x5v, 1x20v); brings us back up to 14 pins. 14's already getting fat enough that it's going to be getting annoying to bend. Adding 4 more pins replace the other half of the 8 pin connector would bump it back to 18; almost as bad as the original 20pin cable and mostly defeating the purpose. With 8 + 4 pin (and occasionally 8 + 8 or * + 4 + 6 (pcie)) additional 12V connectors becoming common on high end mobos; we'd still be connecting multiple cables at the top end.
Going the other direction, only adding a 20v wire to the reduced cable and stopping at 10 pins would make the cable easier to route; but would continue to impose a second power cable on low end mITX boards - the 4pin is ubiquitous even on boards with low power BGA processors and only pcie 1x (25W) slots - where cable routing is most difficult.
I suggested adding 2 12v wires to the redesigned main cable as a reasonable in compromise position. It still gives the big cable a badly needed diet to make routing easier. Low end systems are able to eliminate the 2nd cable entirely. And on the top end, it will at least serve as a counterbalance to the continuing drive for yet another 12v cable; and maybe take a step back on mainstream gaming boxes.
kpb321 - Monday, January 12, 2015 - link
I'm disappointed that it is a single USB c connector as it wastes so much space. In addition to being easier to connect you should be able to fit more ports on the back panel so you don't have 20 usb headers you need to connect to use all the usb ports the board supports.MikhailT - Monday, January 12, 2015 - link
Why are you disappointed? It's too early to replace everything with USB 3.1 right now, there's not enough industry support to do this. The chipsets are not ready, issues with the power delivery, connector designers and so on.It's going to take a few years before USB 3.1 have the proper support from everybody.
MSI is just trying to be the first with USB 3.1 connector and they did it by doing this weird setup with third party chipset, combining ports and so on.
kpb321 - Monday, January 12, 2015 - link
USB C connector does not mean USB 3.1. It is entirely possible to do USB C ports that are USB 3.0 or even USB 2 and there are plenty of USB 2 and 3 ports supported by the chipset for them to not all fit on the back panel.fokka - Tuesday, January 13, 2015 - link
as often with first gen products, oems just wanna "get it out there" first, so one lonely port is enough to slap that "usb3.1" sticker on the box. more ports wouldn't make much sense anyways, if the bandwidth to the chipset is limited. i just can't wait until intel and amd get off their asses and give us glorious usb3.1 integrated into the chipset per default.CrazyElf - Monday, January 12, 2015 - link
I just wish that they would abandon their love affair with Qualcomm NICs. I'd rather see them feature something like Intel server grade NICs on their motherboards - that would be a real upgrade.Otherwise, I'm pretty happy with MSI.
I actually have become fond of their XPower series of motherboards. Their big weak point seems to be RAM overclocking (fewer profiles) and their UI isn't always the best. Oh, and Command Center has been buggy at times.
Their Lightning GPUs are among the best. I just hope they release big Maxwell and whatever AMD offers with top notch VRAM (or whatever HBM equal it is) this time (they had to use Elpida on the 780L, although they later replaced with Hynix).
Otherwise pretty good. Launch day BIOS stability is not good, but they do seem good about releasing BIOS updates and they have good tech support.
Intel currently controls the PCH, and with that, there's limited room for motherboard innovation, but I am hoping to see something interesting in the upcoming years, especially now that single threaded performance doesn't seem to be getting much faster with each generation.
just4U - Monday, January 12, 2015 - link
From the article:"(and get frustrated when it never goes in first time)"
----
Sometimes you don't get a second chance. I've bent the pins on a few boards now.
just4U - Monday, January 12, 2015 - link
Come to think of it.. I could have misread your comment.. I was talking about the 3.0 header on the board that you connect your USB 3.0 case cable to..jdrch - Tuesday, January 13, 2015 - link
Great, so we have to wait for chipset support again as we did with USB 3.0? *sigh*The_Assimilator - Tuesday, January 13, 2015 - link
They put a brand spanking new USB type-C connector on the board... alongside an ancient and obsolete VGA connector.Seriously MSI, what are your engineers smoking?
fokka - Tuesday, January 13, 2015 - link
agreed, vga needs to die!SirPerro - Tuesday, January 13, 2015 - link
With the Type-C being much stronger than the current micro-usb connector, what's exactly the concern for smartphone/tablet makers?fokka - Tuesday, January 13, 2015 - link
i guess it's just cheaper to implement a simple micro usb2.0 port right now and since that's good enough for charging and the occasional data transfer, it will take a while until type-c is ubiquitous on mobile devices.another question is how this will be reglemented in the EU, since right now phones have to come with micro usb and i don't know if you're even allowed to sell a phone with type-c here at the moment.
dabotsonline - Wednesday, January 14, 2015 - link
I believe that Apple supply a Lightning-to-microUSB adapter in European retail boxes so I presume that any manufacturer who wishes to implement USB 3.1 Type-C would do similarly.Also, Ian Cutress, Dong Ngo has reported that even the X99A Gaming 9 ACK motherboard (and GT72 gaming notebook) will come with a Type-C port:
http://www.cnet.com/news/usb-type-c-hands-on-its-h...
http://arstechnica.com/gadgets/2015/01/usb-3-1-and...
Ethos Evoss - Tuesday, January 13, 2015 - link
3.1 better than SATA 6 ... OK SATA you can finis now you useless ..AnnihilatorX - Thursday, January 15, 2015 - link
Do you know that SATA Express and M.2 are SATA 3.2 specifications which uses PCI Express lanes, and therefore can support up to 16Gb/s?eanazag - Tuesday, January 13, 2015 - link
I've liked the look of MSI's motherboards. I have a H55 socket 1156 still running for me with a Core i3.I have two of their Twin Frozr R9 290 cards. I have been pretty happy with them. The fans on them are not 24/7 high rpm and will freeze up after spurting a little oil from the fan internals. I've read online that they can have mineral oil added to them and work as new at that point.
I handled this through warranty. It cost me $18 to ship. MSI's policy says it can take up to 28 days in RMA, but I got mine back in a week, but it wasn't the same card. The card wasn't as new as mine and the part that screws to the case was bent. I had to bend it back to get it to mount into the board and case. I had RMAed a Gigabyte GTX660 and was without the card for about 17 days.
DanNeely - Tuesday, January 13, 2015 - link
With the exception of laptop/desktop computers (where the data on the drive matters); you almost never get the same item you sent in for RMA back. To minimize the turnaround they send you a replacement off their shelves immediately and then shelve yours after fixing it (or determining it's irreparable and disposing of it).I'm mildly disappointed that they made you send the whole card in for a round trip though. A few years back I had a fan die on my XFX(?) 5870. I filed an RMA stating that the fan refused to spin and the card was overheating. They asked if I felt comfortable unscrewing the existing heat sink and replacing it myself. When I said yes, they just mailed me a replacement and didn't require me to return the original.
Aslan7 - Wednesday, January 14, 2015 - link
How is it difficult to get 100 watts to each and every USB motherboard port? Currently pcie x16 slots are required to supply 75 watts each. Many motherboards now have 3-6 of these slots. If you need more power you just plug in another power connector. The biggest difficulty I can see is OEM systems using within 50 watts of the max power output of the power supply. Only in that event would I see it being a difficulty. If it's too difficult to get that much power to USB on the motherboard backplate designers could simply have USB headers that are cabled to a bracket that would fit in the back of a computer and a place on the inside of the bracket to plug in a few power connectors. I predict this will lead to a need for a way for a power supply to communicate with the computer itself to state how much power it has available.AnnihilatorX - Thursday, January 15, 2015 - link
Why do you need every port toi be 100W capable? Realistically only 1 or 2 ports will need such power, and if the motherboard has to provide 100W interface to say 12USB 3.0 ports in the future, the cost involved will be prohibitory. And normal power supplies will not be able to handle even 6 USB3.0 ports each drawing 100W. So what is the point of wiring up all USB ports for power when 1) you only need a few 2) power supply limits?Haravikk - Friday, January 16, 2015 - link
What's with these open-air case designs? Viewing windows are fine I guess, but has no-one heard of dust in the computer-case design world?