For a "regular" USB C that supports USB 2.0 speeds (and is rated for 60W and therefore lacks an internal e-marker chip), there's just 5 wires inside: Two for data, two for power, and one for CC. There's nothing particularly complex about testing those wires for end-to-end continuity (like a cheapo network cable tester does).
A charging-only cable requires only 3 wires.
But fancier cables bring fancier functions. Do you want to test if the cable supports USB 3? With one lane, or two lanes? USB 4? Or what of the extra bits supporting alt modes like DisplayPort and MHL and the bag of chips that is Thunderbolt -- does that need all tested, too? (And no, that earlier 120Gbps figure isn't a lie.)
And power? We're able to put up to -- what -- 240W through some of these cables, right? That's a beefy bit of heat to dissipate, and those cables come with smarts inside of them that need negotiated with.
I agree that even at extremes, it's still somewhere within the realm of some appropriate FPGA bits or maybe a custom ASIC, careful circuit layout, a big resistor, and a power supply. And with enough clones from the clone factories beating eachother up on pricing, it might only cost small hundreds of dollars to buy.
So then what? You test the fancy USB-C ThunderBolt cable with the expensive tester, and pack it up for a trip for an important demo -- completely assured of its present performance. And when you get there, it doesn't work anyway.
But the demo must proceed.
So you find a backup cable somewhere (hopefully you thought to bring one yourself, because everyone around you is going to be confused about whatever it is that makes your "phone charger" such a unique and special snowflake that the ones they're trying to hand to you cannot ever be made to work), plug that backup in like anyone else would even if they'd never heard the term "cable tester," and carry on.
The tester, meanwhile? It's back at home, where it hasn't really done anything but cost money and provide some assurances that turned out to be false.
So the market is limited, the clone factories will thus never ramp up, and the tester no longer hypothetically costs only hundreds of dollars. It's right back up into the multiple-$k range like the pricing for other low-volume boutique test gear is.
(I still want one anyway, but I've got more practical things to spend money on...like a second cable to use for when the first one inevitably starts acting shitty.)
>(And no, that earlier 120Gbps figure isn't a lie.)
Is that 120 Gbps or 120 GB/s as the previous poster stated? 120 GB/s is on the order of DDR5 throughput - I doubt we have any kind of cheap cable tech that can carry that kind of bandwidth right now - while 120 Gbps is more like NVME 5 SSD speed.
Good catch. I didn't notice the factorization difference.
I myself definitely meant gigabits per second. 120Gbps is about what 2x 8k 60Hz monitors use at a constant rate in the simplest sense (pixels), so that's what I assumed they were talking about.
But it's more complex than that.
Looking a bit deeper: It seems that TB5 (which uses USB C) is natively a symmetric 80Gbps: 80Gbps one way, and 80Gbps the other way. 160Gbps, total, counting both directions -- plus or minus overhead.
It can also use a "turbo" mode where things are shifted to be 120Gbps one way (host-to-device) and 40Gbps the other way (device-to-host), which is still 160Gbps in aggregate.
It isn't clear how that dual-8k-display mode provides any extra bandwidth for other things that may be downstream like storage devices, but that's as deep as I feel like going.
And that is deep enough to answer your question: It is definitely in the neighborhood of [up to] 120Gbps [in one direction], and it is definitely never in the neighborhood of 120GBps [neither in any direction, nor in aggregate].
For a "regular" USB C that supports USB 2.0 speeds (and is rated for 60W and therefore lacks an internal e-marker chip), there's just 5 wires inside: Two for data, two for power, and one for CC. There's nothing particularly complex about testing those wires for end-to-end continuity (like a cheapo network cable tester does).
A charging-only cable requires only 3 wires.
But fancier cables bring fancier functions. Do you want to test if the cable supports USB 3? With one lane, or two lanes? USB 4? Or what of the extra bits supporting alt modes like DisplayPort and MHL and the bag of chips that is Thunderbolt -- does that need all tested, too? (And no, that earlier 120Gbps figure isn't a lie.)
And power? We're able to put up to -- what -- 240W through some of these cables, right? That's a beefy bit of heat to dissipate, and those cables come with smarts inside of them that need negotiated with.
I agree that even at extremes, it's still somewhere within the realm of some appropriate FPGA bits or maybe a custom ASIC, careful circuit layout, a big resistor, and a power supply. And with enough clones from the clone factories beating eachother up on pricing, it might only cost small hundreds of dollars to buy.
So then what? You test the fancy USB-C ThunderBolt cable with the expensive tester, and pack it up for a trip for an important demo -- completely assured of its present performance. And when you get there, it doesn't work anyway.
But the demo must proceed.
So you find a backup cable somewhere (hopefully you thought to bring one yourself, because everyone around you is going to be confused about whatever it is that makes your "phone charger" such a unique and special snowflake that the ones they're trying to hand to you cannot ever be made to work), plug that backup in like anyone else would even if they'd never heard the term "cable tester," and carry on.
The tester, meanwhile? It's back at home, where it hasn't really done anything but cost money and provide some assurances that turned out to be false.
So the market is limited, the clone factories will thus never ramp up, and the tester no longer hypothetically costs only hundreds of dollars. It's right back up into the multiple-$k range like the pricing for other low-volume boutique test gear is.
(I still want one anyway, but I've got more practical things to spend money on...like a second cable to use for when the first one inevitably starts acting shitty.)