As usual, TPU doesn't publish the most important metric: random 4k reads. Looking at the included graph, at QD1 it looks like this drive can do ~30k IOPS which would translate to over 100MB/s. Is that right? Because if it is, it's pretty amazing. I mean, still seriously lagging behind other numbers, but seriously impressive compared to every other SSD I know.
Frankly I don't find any of the metrics particularly valuable. I mean if two items are the same cost, fine take the faster one.... but gotta wonder if all the time saved in its life is more of less than time saved picking which on to buy. That's no knock on the article or the author but how many of the tests are relative to things that people a) actually relative to something most users do each and every day. Every once and a while, say even every other year, I'd love to see a test that says ... here's the measured ROI on adding an SSD.... for average PC enthusiast.
Windows Start Up - Best / Worst = 17.5 / 19.0 (I'm on Win 7 but for me when I built it was 15.6 for my old Samsung pro and 16.5 for SSHD to Password Window). But in any case, it's gotta be said, why do we care ? Will anyone accomplish anything in that 1.5 seconds time. The automatic response is Well 1.5 seconds x 365 days x 3 years = 1,642.5 seconds ... in other words, 27 minutes. More often that not, you'll spend more than that selecting, unboxing, reading, installing and testing the thing. And the reality is... well I only reboot once every 2 weeks so I'll save < 2 minutes. Did I accomplish anything ?... not really cause whether it's 17 seconds or 60 seconds is irrelevant as while it's booting, I'm listening to phone messages for 5 minutes and calling those folks back for 15 more.
Uncompress WinRAR - Not something I imagine folks do more times than I figure they have fingers on 1 hand in a box's lifetime. But while it's Uncomprressing, I'm alt-tabbed out doing something else so why do I care ?
Importing 1000 MP3 songs into Music Library - Never done that, and if I did, not as if a) Wouldn't waste SSD space on a music library, b) might do it once every 4 years and c) would like to say I'd be be alt-tabbing out to multi-task while it's happening but by the time I exhaust that 0.7 seconds I'd be done.
ISO Copy / Paste - Again, a) Wouldn't waste SSD space on a ISO, b) might do it once every 4 years and c) would be alt-tabbing out to multi-task while it's happening.
MS Office - See ISO
Virus Scan - Happens every day, I don't even know it's happening so no impact
Apple iTunes - a) might do it once every 4 years and b) would like to say I'd be be alt-tabbing out to multi-task while it's happening but by the time I exhaust that 2.1 seconds I'd be done.
Google Chrome - a) might do it once every 4 years and b) would like to say I'd be be alt-tabbing out to multi-task while it's happening but by the time I exhaust that1.6 seconds I'd be done.
Adobe Reader- a) might do it once every 4 years and b) would like to say I'd be be alt-tabbing out to multi-task while it's happening but by the time I exhaust that 2.6 seconds I'd be done.
Adobe PhotoShop - a) might do it once every 4 years and b) would like to say I'd be be alt-tabbing out to multi-task while it's happening but by the time I exhaust that 0.5 seconds I'd be done. Better approach here might be to double click on a 8 MB AutoCAD 2019 file and time how long it takes to start editing.
PhotoShop Editing - here we fall outside the realm of the average PC user and for all CS things Adobe, SSDs are almost an automatic for scratch files. Here we are seeing 10 50 MB images files go thru 4 operations ... looking at each image individually = 0.875 seconds difference per image. What about user time ? No one is going to look at the results ? So is it really thet big a time save if one is looking at Image 2 while 1 is being processed.
Gaming Level Loading - We are looking at 2.9 and 0.4 seconds. Now when I complete a level and i see the fade out, I'm using the break for a bio, snackie, feed the cats, check my scores, let the dog out, unflatten my butt cheeks, straighten my legs, check my phone.
Now I'm as nerdy as the next geek and yes if I was doing a build today, it would likely have more than one 970 in it. What I hate to see however is users pushed to add a SSD at the cost of a GFX card tier. If build w/ 970 Pro SSD and HD or SSHD for bulk storage with a 1080 hits the users budget limit, then dropping the SSD to grab a 1080 Ti with a 33% increase in performance to my eyes is the proverbial no brainer. Anyone will notice 33%. But running dual boot desktops and laptops with SSD + HD/SSD versus SSHD over 6 weeks with 5 users .... no one noticed.
I recall paying $1,000 for a SCSI 1 GB Hard Drive. I don't think it will be along ti me before 2 TB od SSD is only marginally more expensive than a 2 TB SSHD ... when that difference is < $100it will be something no one has to think about.
But getting back to that biennial test. I know that there's no way I could possibly justify the purchase of a high end SSD. But I am going to by 2 of them anyway
. But I would enjoy laughing at myself a bit just to keep it real ... so it would be gteat if we could somehow measure the real daily impact and ROI. If we don't check ourselves a bit, I'll start trying to rationalize 3 new HDR Screens in one of these $6 k gaming chairs.
