I fail to see the appeal of new PCIe.
Sure, we can go on about "muh bandwidth" however, in a consumer setting, bandwidth is not really an issue anymore. There is no difference between PCIe 3.0 x8 and x16 on a single video card, the only thing GPU wise that ever pushed slots to their limit was dual GPU, which is not long dead. And thats on the highest end hardware, which only a small # of people use. The vast majority of PC gamers are still below the 1060 in performance, and play at 1080p or lower. This puts the number of gamers pushing for new PCIe as a niche of a niche market.
What about NVMe you ask? Well, I ask, why do you need more? NVMe drives today are easily 6-7X faster then the best SATA III SSDs, and yet the real world results show that does diddly squat. Load times for windows are, what, a whole second faster? Maybe 2 in a pinch? IME, CIV V takes just as long to load from my 950 pro as it did off of my old mushkin reactor SSD. Waiting for windows to load takes almost the same amount of time, and waiting for programs like STEAM and Afterburner to load after launch shows no difference either. Multiple benchmarks show no difference in consumer software between SATA and NVMe. Bulk file movements, like for 4K video editing benefit, but only on a local machine, and transferring between devices is limited by network speeds, and file sizes big enough to benefit make USB 3 drives irrelevant, even then the best only hit around SATA III speed, and super fast high density USB C 10 Gbps drives are unicorns.
I feel that we are hitting the sharp, hard wall of diminishing returns. Modern computer hardware was already way too fast for tasks outside of 3d gaming or productivity work, and this push for stupid fast internal bus speeds seems to be ignoring the potential applications for consumers when network, flash, and external storage already cant fully utilize what we have, and professional markets are already using x16 card SSDs, again asking what the consumer market well get out of this new, expensive tech. I see people here saying that having faster lanes will allow more connectivity, but surely having MORE PCI3 lanes so I dont need to split them between interfaces would accomplish this goal just as well at a far lower cost?
To me, PCI4 is just like USB 6.9 2x2 ultra mcnugget drives. Most people I know still have and use 2.0 USB drives, a few use 3.0, and those are already more then fast enough for the files they are moving. Much like how SATA III SSDs and USB 3 flash drives are more then enough for consumers, I feel PCI3 is already faster then most users will likely utilize in the next decade.