You will notice the increased responsiveness immediately and then gradually become accustomed to it. In our experience, you will take the performance for granted until you work on a slower PC.
Therein lies the rub. *I* certainly notice the improvement with a SSD. But the vast majority of the clients whose HDDs I've replaced with a SSD report little to no "noticeable" improvement. One of them even canceled my recommendation to upgrade all their work computers to SSDs after trying it out in one computer for a couple weeks. Their employees couldn't even tell when they were using a computer with a HDD vs the computer with the SSD.
So the improvement going from a HDD to SSD (which is massive compared to SSD to Optane) isn't that big a deal to most users.
takeshi7 :
It's obvious that once an SSD is installed in a system, the game loading times become CPU and/or RAM bottlenecked, not storage bottlenecked. That's why these fancy NVMe SSDs barely load games faster than budget SATA SSDs.
SSDs are currently constrained by 4k random read/write performance. That's why NVMe SSDs barely load faster - their 4k speeds are nearly the same as for SATA SSDs. See below.
mdd1963 :
Will need to see some more comparisons, but, I was expecting peak numbers to be much better than 2500 MB/sec reads.....; my damn 960 EVO will do 3,200 MB/sec sequential reads...
Despite it being The Metric everyone uses to compare, the sequential speeds actually aren't that important. You see, how fast a drive "feels" depends on how long you have to wait for it to process a request. That is, how many seconds you have to wait. MB/sec is the inverse of that. What you really want to compare is sec/MB. Look what happens when you compare a bunch of speed increases in sec/MB for reading 1 GB.
125 MB/s HDD = 8 seconds
250 MB/s SATA2 SSD = 4 sec
500 MB/s SATA3 SSD = 2 sec
1000 MB/s early PCIe SSD = 1 sec
2000 MB/s NVMe SSD = 0.5 sec
Notice how the reduction in wait time becomes smaller every time you double MB/s. So even though the jump from SATA3 to a NVMe SSD looks like it should be massive (+1500 MB/s), it's only a wait time reduction of 1.5 seconds. Only 25% of the wait time reduction you got upgrading from a HDD to a SATA3 SSD (6 sec reduction in wait time). So the bigger MB/s gets, the less difference it makes in perceived speed (wait time reduction).
Likewise, the sequential read/writes happen so quickly that they add only a small amount to your wait time. The biggest contributor to wait time is the
slowest operation of the drive. For both HDDs and SSDs, this is the 4k random read/writes. About 1 MB/s for HDDs, and 30-70 MB/s for SSDs. Even the NVMe SSDs are struggling to break 100 MB/s at 4k writes (reads are slower). If you're writing 1 GB of sequential data and 1 GB of 4k random data, the 2000 MB/s NVMe drive will finish the sequential write in 0.5 seconds, while taking approx 20 sec to do the 4k write (QD=1).
So it's the
slowest operation of the drive which matters the most, and Optane appears to deliver more than a 5x improvement in the slowest drive operation. The only question that remains is whether that difference will be enough to show up when inverted to sec/MB. According to the graph, the answer would appear to be no. If you're doing a 100 MB 4k read (QD=1):
0.94 MB/s HDD = 106 seconds
38.98 MB/s SATA SSD = 2.6 sec
44.5 MB/s NVMe SSD = 2.2 sec
251.9 MB/s Optane = 0.4 sec
So the jump from a SSD to Optane (2.2 sec) only gives you 2% additional wait time reduction compared to switching from a HDD to a SSD (103.4 sec). The casual user upgrading from a HDD isn't going to be able to tell the difference between a SATA SSD, a NVMe SSD, and Optane. OTOH an enthusiast used to running with a SSD should notice the difference switching to Optane.
Facebook
Twitter
Instagram