Then there's optane that got ~10us with. The newest controllers and nand is inching closer with randoms but optane is still the most miraculous ssd tech that's normally obtainable
Eventually we'll have machines with unified memory+storage. You'll certainly have to take a bit of a performance hit in certain scenarios but also think about the load time improvements. If you store video game files in the same format they'd be needed at runtime you could be at the main menu in under a second.
The separation into RAM and external storage (floppy disks, magnetic tapes, hard drives and later SSD etc) is the sole consequence of technology not being advanced enough at the time to store all of the data in memory.
Virtual memory subsystems in operating systems of the last 40+ years pretty much do exactly that – they essentially emulate infinite RAM that spills over onto the external storage that backs it up.
Prosumer grade laptops are already easily available, and in 2-3 years there will be ones with 256-512 Gb as well, so… it is not entirely incoceivable that in 10-20 years (maybe more, maybe less) the Optane style memory is going to make a comeback and laptops/desktops will come with just memory, and the separation into RAM and external storage will finally cease to exist.
P.S. RAM has become so cheap and has reached such large capacity that the current generation of young engineers don't event know what a swap is, and why they might want to configure it.
I am also of the opinion that we are heading towards the convergence, although it is not very clear yet what the designs are going to converge on.
Pretty much every modern CPU is a hybrid design (either modified Harvard or von Neumann), and then there is SoC, as you have rightfully pointed out, which is usually modified Harvard, with heterogenuous computing, integrated SIMT (GPU), DSP's and various accelerators (e.g. NPU) all connected via high-speed interconnects. Apple has added unified memory, and there have rumours that with the advent M5 they are going to change how the memory chips are packaged (added to the SoC), which might (or might not) lay a path for the unification of RAM and storage in the future. It is going to be an interesting time.
Memory and compute tend to use different processes and putting everything on the same wafer is also bad for yields - you only do it if there is a really good reason.
The storage layer isn't just there for swap - that's not even its main purpose. The need for persistent storage is not going to go away and if you can't cleanly separate that from your runtime state then "have you tried turning it off an on again" now becomes "have fun reinstalling the whole thing".
At a minimum, we should be able to get everything to DRAM speeds. Beyond that you start to run into certain limitations. Achieving L1 latency is physically impossible if the storage element is more than a few inches away from the CPU.
This is what makes PS5 and Xbox so crazy. Loads for stages are 2-5 seconds and some have optimized it for no loading on open worlds. PS4 and below took forever to load sometimes. Like 30 seconds, which is insanely long.
That would be possible even on spinning harddrives as long as they're already spinning.
The fastest memory can't prevent the real reason games take so long to the menu: company logos. The Xbox can already resume a closed game within a few seconds, loading a simple main menu is trivial in comparison.
I see an order of magnitude higher prices than 990 PRO / SN850X (and almost 9100 PRO / SN8100). While Optane seems to perform "only" 4 times faster in 4K Q1T1 (read) and 3-4 times faster overwriting the whole drive as quickly as possible (any use case for the later besides restoring from backup?). It looks like in less synthetic benchmarks the difference becomes more significant (only for the very last Optane P5800X).
