[SOLVED] USB 3.2 gen 2x2 to Thunderbolt 3 adapter?

[armpakorn]

I have just bought SanDisk Extreme Pro Portable V2 2TB SSD which claims to support USB 3.2 gen 2x2 for its maximum read/write speed of 2000 Mb/s. Unfortunately, my laptop computer doesn't support USB 3.2 gen 2x2 but supports Thunderbolt 3 (which has a bandwidth of 40 Gbps so I think that can accommodate the USB 3.2 gen 2x2 bandwidth of 20 Gbps)

Is there such a thing like "USB 3.2 gen 2x2 to Thunderbolt 3 adapter". I tried searching but fruitless.

 

[Lutfij]

Welcome to the forums, newcomer!

Nope, there isn't.

 

[armpakorn]

Is it possible to use a PCIe to Thunderbolt 3 adapter (something like this: here) and then connect an USB 3.2 gen 2x2 card (Example) and use it as an adapter?

 

[toaster42]

Is it possible to use a PCIe to Thunderbolt 3 adapter (something like this: here) and then connect an USB 3.2 gen 2x2 card (Example) and use it as an adapter?

Yes, that could work.
In reality the primary use case of such thunderbolt 3 connected PCIe enclosures is to use a PCIe graphics card externally (aka eGPU, generally for laptops, though it could also upgrade a desktop system, I suppose.). I'm mentioning the use case because when this product category first released it was plagued with driver issues. Many/most were fixed, but with an eye towards the primary use case.

My gut tells me that it either will not work, or at least not be stable enough to use on a drive that hasn't been thoroughly backed up. Additionally, the cost of an eGPU enclosure plus USB 3.2 gen 2x2 card will likely exceed the MSRP of your SSD at $350. Adapters which double the investment in a technology generally do not fare well in the consumer (or if we're being honest, any) market.

 

[toaster42]

Actually, I just noticed the nvme PCIe card they optionally ship that 'eGPU' chassis with and realized you're intended to use it in the external PCIe chassis, not your desktop, as I had originally imagined. Clearly they are confident in that external PCIe chassis' use with nvme. A SATA card just has an additional protocol translation step. Therefore, I will slightly revise my earlier assessment. While I still find it unreasonably expensive and deem it unlikely to work well, I can also see how it is plausible that it may actually work (albeit with some sort of performance penalty for the extra translation step in the pipeline)