Thursday, August 29th 2024

ASMedia ASM4242 USB4 Host Controller Chip Achieves Thunderbolt 4 Certification

ASMedia announced that its USB4 host controller chip, ASM4242, has received Thunderbolt 4 certification, following its USB-IF certification earlier this year. This marks a significant milestone for the ASM4242, toward providing a comprehensive high-speed transmission solution and an outstanding user experience. The Thunderbolt 4 certification signifies that ASM4242 meets Thunderbolt's stringent standards in signal quality, compatibility, and reliability, ensuring exceptional connection stability and compatibility.

Thunderbolt 4 is fully compatible with USB4, integrating power, data transmission, and video display into a single interface, offering up to 40 Gbps bandwidth to meet the demands of high-speed file transfer, 100 W fast charging, and multi-display setups. A single cable can daisy-chain up to 5 devices, driving high-bandwidth data, power, and video transmission. Gamers, creators, and professionals can connect multiple displays and accessories through this powerful connection standard. They can also enjoy a clean and convenient workspace.
ASM4242 is backward compatible with previous Thunderbolt and USB versions, and also supports DisplayPort and PCIe standards, making it suitable for a wide range of applications including laptops and desktops. The versatility of ASM4242 allows for flexible integration into various high-end consumer electronics, providing users with faster and more convenient connectivity.
Add your own comment

8 Comments on ASMedia ASM4242 USB4 Host Controller Chip Achieves Thunderbolt 4 Certification

#1
TheLostSwede
News Editor
I don't understand why they used a fake image of the chip.
This is what it looks like.
Posted on Reply
#2
Cheeseball
Not a Potato
Nice. Hopefully it will be used more than Intel's old Maple Ridge crap (JHL8540) now that its certified.
Posted on Reply
#3
Assimilator
Fucking finally, I was honestly expecting Intel to announce Thunderbolt 6 before ASMedia got ASM4242 certified and shipped (this is also a dig at Intel because a year after their TB5 announcement there are still almost no TB5 devices).

And yes, before people remind me that it's Intel's fault for holding up the certification... it's partly Intel's fault. It's also partly AMD's fault. But it's mostly ASMedia's fault IMO, and I don't think it's unfair to criticise them when this functionality was originally supposed to launch with Zen 4, two whole years ago which is basically an eon in technology terms.

Also I just realised that TB5 will need 12 lanes of PCIe 4.0 which is basically an entire x16 slot, meanwhile AMD and Intel are gaslighting us with a single x16 slot's worth of IO connectivity. How 'bout you mofos move some of those PCIe lanes down from "HEDT" (actually server now because it's so overpriced) to consumer CPUs? Or heaven forbid, make consumer chipsets that aren't effectively worthless (the TRX40 that I am using to type this message on is from Zen 2 days, yet puts out more PCIe 4.0 lanes than AMD's "best" consumer-grade Zen 5 X870E offering - how pathetic).
Posted on Reply
#4
TheLostSwede
News Editor
AssimilatorFucking finally, I was honestly expecting Intel to announce Thunderbolt 6 before ASMedia got ASM4242 certified and shipped (this is also a dig at Intel because a year after their TB5 announcement there are still almost no TB5 devices).

And yes, before people remind me that it's Intel's fault for holding up the certification... it's partly Intel's fault. It's also partly AMD's fault. But it's mostly ASMedia's fault IMO, and I don't think it's unfair to criticise them when this functionality was originally supposed to launch with Zen 4, two whole years ago which is basically an eon in technology terms.

Also I just realised that TB5 will need 12 lanes of PCIe 4.0 which is basically an entire x16 slot, meanwhile AMD and Intel are gaslighting us with a single x16 slot's worth of IO connectivity. How 'bout you mofos move some of those PCIe lanes down from "HEDT" (actually server now because it's so overpriced) to consumer CPUs? Or heaven forbid, make consumer chipsets that aren't effectively worthless (the TRX40 that I am using to type this message on is from Zen 2 days, yet puts out more PCIe 4.0 lanes than AMD's "best" consumer-grade Zen 5 X870E offering - how pathetic).
You're aware that the ASM4242 has been available since last year and has been certified for USB4 for over a year by now, right?
This is just an additional certification to show that it's Thunderbolt 4 compatible.
MSI has had this for sale for close to a year.
www.msi.com/PC-Component/USB4-PD100W-EXPANSION-CARD
Posted on Reply
#5
Assimilator
TheLostSwedeYou're aware that the ASM4242 has been available since last year and has been certified for USB4 for over a year by now, right?
This is just an additional certification to show that it's Thunderbolt 4 compatible.
MSI has had this for sale for close to a year.
www.msi.com/PC-Component/USB4-PD100W-EXPANSION-CARD
Yeah, and it has an MSI-proprietary GPIO header which means you can't use it unless you have a MSI motherboard, which means it immediately falls into the exact same hole of uselessness as the add-on TB3 cards that had a proprietary GPIO header. I don't understand why these cards need GPIO that can't be provided by the PCIe bus, and I don't understand why USB-IF can't define a standard for this if it is indeed necessary.
Posted on Reply
#6
TheLostSwede
News Editor
AssimilatorYeah, and it has an MSI-proprietary GPIO header which means you can't use it unless you have a MSI motherboard, which means it immediately falls into the exact same hole of uselessness as the add-on TB3 cards that had a proprietary GPIO header. I don't understand why these cards need GPIO that can't be provided by the PCIe bus, and I don't understand why USB-IF can't define a standard for this if it is indeed necessary.
It's not just GPIO though, but it's still not public what goes into that connector and yes, all the motherboard makers have their own connector for their Thunderbolt/USB4 cards, which is just plain stupid.

The card still work without it, but hot plugging doesn't behave properly, nor does sleep and you can't do video through afaik either. But you do get Thunderbolt/USB4 data without it...
So does that at least explain why it's needed?
But yes, it should be a common standard.
Posted on Reply
#7
Assimilator
TheLostSwedeSo does that at least explain why it's needed?
I don't understand what's so special about what these cards are doing that they can't do it through the PCIe bus. And if there is something, why can't the PCIe and/or USB specs be amended to handle it, instead of this extra-proprietary-cable nonsense?
Posted on Reply
#8
TheLostSwede
News Editor
AssimilatorI don't understand what's so special about what these cards are doing that they can't do it through the PCIe bus. And if there is something, why can't the PCIe and/or USB specs be amended to handle it, instead of this extra-proprietary-cable nonsense?
As you've already said, it's a GPIO interface, which can't be done over PCIe. It also appears to be I2C and UART according to the ASM4242 specs.
The interfaces are used for communicating things like hot plugging and sleep modes, as mentioned, which again can't be done over PCIe.
I doubt the PCI-SIG would want to make changes to the PCIe spec to handle some rare case add-in cards for what in many ways is a competing spec.
Due to these standards organisations not talking to each other, we have the mess we have.
In all fairness though, the GPIO + whatever cable could be common among all the motherboard makers, but they decided to be annoying and all make their own.

Also note that USB4/Thunderbolt cards need a USB 2.0 cable, in addition to the other cable, due to not having native USB 2.x or 1.x support, but if you're not planning on using older USB stuff, that can be skipped. On top of that, Thunderbolt doesn't support 20 Gbps USB 3.2 Gen 2x2...

Forgot that Asus has a card too. Same issues though.
www.asus.com/motherboards-components/motherboards/accessories/usb4-pcie-gen4-card/
Posted on Reply
Nov 12th, 2024 16:58 EST change timezone

New Forum Posts

Popular Reviews

Controversial News Posts