They absolutely do!
Certified cables do not.
And there we have it.
and dCS wants us to use certified cables
I think most manufacturers would say the same, they welcome the use of aftermarket cables, but only those that conform to relevant industry standards. There is just so much funky stuff out there, both legitimate and complete voodoo, that no other position is tenable.
I have a DIY ethernet cable for the last hop from my EtherRegen to my Vivaldi. Coincidentally it is a CAT 7 cable with the shield lifted at one end just as described in the Alpha Audio article (so it doesn’t conform to IEEE 802.3) but it works absolutely fine.
So I interpret it to mean that only certified cables are supported. Use whatever makes you happy (as long as it is safe and doesn’t damage your equipment) but if it doesn’t conform to standards and it doesn’t work you’re on your own.
I was sure they didn’t and my mate who works in IT and networking was also sure. As a new Vivaldi stack owner I borrowed from my brother a spare Audioquest Diamond cable for the last connection from my Cisco switch/optical break.
Yes, they do seem to make a difference. We both heard it, leading to much head scratching.
The consensus in this thread so far seems to be along the lines that only properly certified cables may not exhibit sonic changes. So is the AQ Diamond properly certified?
DCS also state that the ethernet cable connecting the dCS device to LAN should be unshielded. AFAIK AQ Diamond is shielded.
Given that neither of the dCS provisions may be met then this might result in some change to the sonics. Just a possibility.
Yes the AQ Diamond is shielded. Unfortunately I hear the change wrought by the AQ to be an obvious improvement. Bit of a bugger really 
This is a risky post. @Anupc may lose his mind! (!)
; )
I run inexpensive, spec, unsheilded ethernet cables, consistent with dCS’s prior written recommendation on this forum.
nah. Instead of measuring the actual digital or analog output, these AA guys are intentionally measuring the wrong thing, and attempting to infer correlation. No need to waste my time trying to rebut.
Hi @keiserrg,
I also use no-frills flat unshielded CAT6A cheap-as-chips cables myself at home…
Sounds like an opportunity for a cable company to send you something to try.
Another fun piece from our friends at Alpha Audio:
This one focusing on the “sound” of software (runs for cover) and different factors that can affect it.
They post about all the sophisticated testing environment and equipment they have etc… and then they drop a nugget like this;
Audirvana converts the incoming stream, usually in a compressed format such as FLAC, MP3, AAC or Apple Losless, into RAW PCM. RAW PCM is the format that a DAC converts to analogue sound.
Our hypothesis is: the conversion is done in memory and in the CPU cycles and, as with all conversions, jitter can be introduced into the signal by the electrical activity of the computer. Since jitter is then already in the signal, the Motivo cannot filter it out at the ethernet port, the electrical connection is not the source of possible distortions.
And it becomes painfully apparent that these guys don’t seem to have any technical or engineering background whatsoever (keep in mind the context is a stream or local FLAC/etc file converted to PCM streamed over UPnP) 
Pseudo-science at its finest
Hi, I just auditioned a $5,000 Siltech Double Crown ethernet cable. Vs my Transparent $300 ethernet. I have the Siltech on order. Whole different experience with Siltech…off my NAS drive into my upsampler…
I have used the audioquest diamond for many years on my rossini, and vivaldi.
It’s a very nice sounding cable i found, can be slightly lively sounding over natural.
I tried a few different cables and all sounded different.
I ended up with the shunyata omega