SATA SSD vs NVMe SSD

[Fuzzyboy]

I'm building a new system and I'm torn between the 500/512 GB versions of the Samsung 950 Pro and 850 EVO.

Do any of you have experiences with the two that you'd care to share? Will the 950 Pro significantly improve load times? (in relation to 5-boxing WoW from 1 PC)

[Xeneres]

I'm building a new system and I'm torn between the 500/512 GB versions of the Samsung 950 Pro and 850 EVO.

Do any of you have experiences with the two that you'd care to share? Will the 950 Pro significantly improve load times? (in relation to 5-boxing WoW from 1 PC)

My work computer has a raid 0 950 pro m.2 ssd and it caps out at about 3.3 GB/s. I also have an 850 pro raid 0 ssd raid 0 for comparison in the same computer. I can definately feel the speed of these two 950s in raid 0 vs my SSD raid 0 as its about 3x faster and can handle more IOPS at once.

Samsung says the 950 m2 ssd can do about 2.5 GB/S read speed and 1.5 GB/s write speed with just one 950 pro. You can also do much more IOPS than a regular SSD. I believe the majority of the bottleneck lies in how my Asus z170 deluxe handles/shares nvme bandwidth not allowing it to get the expected 4-5 GB/s from the 950 pro m2 ssd raid 0.

My SSD raid 0 peaks at about 1.1 GB/s or around 500-550 MB/s read speed with just one drive.

I would imagine it would cut down your load times. One 950 pro is theoretically about 4-5x faster read speed than one 850 pro Sata SSD. I believe the evos are slightly slower and have a slighty smaller capacity at ~ 500 gb vs ~512 vs the pros.

My home computer uses 2x 512 gb 850 pros in raid 0 and they are pretty snappy at load screens with my 5 clients. I also run my OS from this 850 SSD array so that may slow it down its performance slightly

When I get around to upgrading my computer in the next year or so I'll upgrade to m2 and NvME drives

[Ughmahedhurtz]

I have both the SATA m.2 and the newer PCI-E m.2 and the PCI-E is 2-2.5x that of the SATA interface.

AttoBench shows the SATA model showed about 500MB/s r/w, whereas the PCI-E shows about 1GB/s write and 2.2GB/s read. Yes, it's noticeable. The PCI-E in RAID 0 is just ludicrous.

[MiRai]

I have to disagree with those above me when it comes to perception for normal everyday tasks, including gaming. I've been using SSDs (in non-RAID) for probably 6 years now, and haven't had a magnetic disk in my main system for the majority of those, so the moment something changes in the read/write speeds I would imagine I'd notice it. I'll also point out that something like gaming generally relies on 4KB random reads, where all SSDs perform very similarly. Sequential R/W is mostly used for larger files which are read from start to finish, sequentially. Smaller files are generally stored in random places throughout the drive, and also accessed randomly by the game depending on what's happening.

With that said, I've multiboxed from a RAM drive with sequential read/write speeds upwards of 8GB/s and random R/W of ~1GB/s, but I never noticed anything different while playing a game that made me say, "Oh, wow... that loads a lot faster now." (my first and second conclusions from that thread)

Could something have changed since I wrote those conclusions a few years ago? Sure, and maybe I should re-visit it since games do have higher quality textures these days (which are larger in size) and we're no longer using MPQ files (for World of Warcraft), but a single SSD will already net you mere seconds of total loading time, so there really isn't much you can shave off of that since online gaming is also dependent on both latency and internet speed—neither of which an SSD can affect.

Today, I'd only recommend an m.2 drive if you just want the convenience of having a drive mounted right on your motherboard to cut down on the two additional cables (power/data) you won't have to run if you didn't have one. I have a 512GB 950 Pro which I use for my main system drive, but it's only for the convenience I mentioned above, as well as just wanting to finally delve into the world of m.2 drives to see what they were all about.

I also wouldn't throw any heavy tasks at an m.2 drive unless it was properly cooled by a heatsink, since they can get hot and then throttle, effectively making them not all that much "faster" than your standard SSD. However, to be fair, it'd probably require a large file transfer or something like recording or rendering video to the drive in order to reach that point of throttle.

[Ughmahedhurtz]

There are three places I notice the speed differences: booting/rebooting, launching games, and transitioning between levels within a game. Boot times were significant, on the order of 12 seconds after POST to the desktop versus 20 seconds; not world-shattering but definitely noticeable and only moreso as you load down a system with drivers and apps, etc. Game launch differences were pretty similar but there's so much unpacking/decompressing/malloc'ing going on then that it masks some of that resulting in gains that may or may not be subjectively noticeable if you're busy grabbing a drink or fiddling with something else. Where I really noticed the difference was in level transitions in games like Half-Life Ep. 2 or Supreme Commander, and to a lesser extent zoning into Orgrimmar. The thing is, it tends to be cumulative and while you might not notice your load times going from a minute to 2 seconds, my experience was that the overall quality of life was well worth the upgrade.

That said, once you pass the 1GB/sec boundary, I presume the bottlenecks are now memory and interconnects -- I didn't really notice a lot of difference between two 840 Pro SATA 2.5" models in RAID 0 (usually gave me 1-1.2GB/s r/w) versus a single 950 PCI-E m.2 except in random writes where the m.2 would stall on me. Moving to two 950 Pros was very noticeable again, especially in writes. Copying a 3.4GB Linux ISO around in a few seconds just gives me a warm fuzzy.

In hindsight, it would have been much less hassle in terms of UEFI install shenanigans (and probably cheaper at the time) to have just picked up 3x 850 EVO SATA 2.5" drives and run them in RAID 0. That would have put me at ~1.5GB/s read and write versus 1.5r/2.4w on the PCI-Es, which would have been much less of a difference. I'd do some measurements for you but honestly, I am just unmotivated to go tearing apart my drives again.