alcoholbob
Diamond Member
- May 24, 2005
- 6,369
- 437
- 126
I'm with @BFG10K here. Even on desktop Windows applications, I barely notice any difference between my two Ryzen R5 3600 rigs, between the 4.2Ghz all-core OC with RAID-0 NVMe (2x QLC Intel 660p 1TB), and the one running stock with a 240GB Adata SP550 TLC SATA SSD.
I'm sure that I could find a workload that was faster on the RAID-0 NVMe, like Windows Defender all-drive scanning (probably, haven't timed it, the RAID-0 build has more files though, so it might actual take long, in real time, even though it might go faster). But for most things, no noticeable difference.
But isn't this just a matter of the application specific optimization? There's no incentive for anyone to optimize based on the disk speed because PC hardware is all over the place. With the next console generation, games will be specifically tweaked to exploit the new hardware to maximize performance. I'm not saying necessarily current nvme drives will benefit from this, but perhaps some next gen ones will, and maybe only accidentally. I do think PC ports will be lazy in general, but sometimes having similar hardware helps. A good example is Detroit Become Human is a pretty lazy PC port that expects 8 core CPUs because of the consoles. It runs like an absolute dog on anything under 8 cores with 100% CPU utilization, regardless of clock speed or IPC of the CPU itself.
Either way, I still like NVME drives for their compactness. They really help with ITX builds as you can add more storage without taking up much real estate in the case, regardless of whether they are faster or not.