How fast is your hard drive?
250GB Samsung 850 EVO SSD
>>55447667
>your hard drive
>posts SSD
wew lad
>>55447667
You turned on Rapid Mode, didn't you?
>SSD
>>55447694
What? I didn't mess with settings on my SSD as I can't on Linux.
>>55447694
yes, he did. sata ssd's are capped at 550mb read speed
pic related is an actual NVMe drive
>>55447713
http://www.anandtech.com/show/9023/the-samsung-ssd-850-evo-msata-m2-review
delete yourself
reminder that ssd's will be deprecated within a year
https://www-ssl.intel.com/content/www/us/en/architecture-and-technology/3d-xpoint-technology-animation.html
https://en.wikipedia.org/wiki/3D_XPoint
http://www.zdnet.com/article/how-intels-3d-xpoint-will-change-servers-and-storage/
>>55447750
What the fuck? I just pressed "All" when I opened up the application. So those are my numbers m8.
SM951 nvme. Fast.
>>55447667
Samsung SM951 M.2 here
>>55447805
In Samsung's software there is a 'rapid mode'. It just caches stuff to ram which is very fast
>>55447805
good job managing to beat the specifications which clearly state the 850 evo is capped at 540 mb read speed. you clearly do not know what you're talking about, yet another tripcode to filter.
>>55447784
>almost dram speeds
>for storage
based as fuck tbqh
just put the pagefile on it and you'll never need more than 16GiB of ram
>>55447841
I didn't enable it though. I just popped the drive in my laptop and installed Linux on it, so any of the Samsung Magician Windows only software won't work for me.
>>55447828
looks like mine >>55447750 is throttling like hell
Not bad for $38.
>>55447709
>this image shows that hdd is faster than ssd
>ssd is faster in reality
Simmer down anon your autism is showing
>>55447709
I understand this is a troll image, but even for that, it's a particularly fucking retarded troll image.
The NAND is accessed in parallel, not in sequence.
I wish I could get in on the SSD meme, but I am but a broke NEET. Plus, I have no free SATA ports.
>>55447859
How do you think your drive can achieve beyond SATA 3 speeds, retard? Just because you're not using rapid mode doesn't mean your OS isn't caching natively.
>>55449118
>>55448032
>less than 2k read
AHHAHAHHAHAHAHHAHAHAHHAHAHAHAHHAHAHAHHAAHHAHHAHAAHHAHAHHAHAHHAHAHAHAHHAHAHAHHAHAHAHAHAHHAHAHA
kill yourself poorfag
>>55447667
>>55447667
120 GB Samsung 850 Pro SSD
>>55449420
Nobody cares what Rapid Mode speed does, you're basically using ram instead of the actual SSD for the test.
No shit ram is faster, why do you think people actually set up RAMdisks?
>>55449518
>Rapid Mode
I'm not using Rapid Mode. I'm also not using a RAM write cache for this test.
These are direct speeds to the hardware. I can tell you why they're so fast, but that would ruin the fun.
>>55449533
Can you tell me why they're so fast? :)
>>55449570
It's mostly because CrystalDiskShit is a shitty benchmark.
It tests throughput by just writing a bunch of 0s. Since I'm using btrfs with LZO compression enabled for my /, this result in tiny files on disk - and therefore CDM thinks it's blazing fast, basically bounded by how fast it can compress/decompress LZO.
In a real benchmark (e.g. fio or iozone) that isn't some GUI winbabby shit, you would test with random data to combat invalidating the results like this on modern filesystems (btrfs, zfs, etc.).
I guess my point is twofold:
1. CrystalDiskMark is barely a benchmark. Really, use fio!
2. Filesystem compression exists and it *really* helps for stuff like log files, source code, text documents and other highly compressible content.
>>55449672
https://0x0.st/Ag2.txt
Here's a FIO result for comparison.
>>55449672
Thank you!
>>55449570
You are truly an idiot.
>>55449672
You, sir, are the exact opposite.
>>55449707
With libaio instead (slightly better)
https://0x0.st/Ag_.txt
>>55449758
Okay
