• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

What is your AS SSD Benchmark score?

HyperX predator M.2 boot drive and a Samsung 850 EVO 500gb steam drive
 

Attachments

  • as-ssd-bench Kingston SHPM228 5.9.2016 8-40-51 PM.png
    as-ssd-bench Kingston SHPM228 5.9.2016 8-40-51 PM.png
    29.1 KB · Views: 1,241
  • as-ssd-bench Samsung SSD 850 5.9.2016 8-32-31 PM.png
    as-ssd-bench Samsung SSD 850 5.9.2016 8-32-31 PM.png
    29.5 KB · Views: 1,264
Seg22 said:
Can someone explain how this is possible with Sata3 (6 gbit/s)?
Click to expand...
Samsung Magician's RAPID mode uses your computer's RAM (which is very fast) as a buffer. When you test with a small size (1GB), the test runs in the buffer and hardly touches the drive and therefore reports very fast speeds. When you test with a large size (10GB), it overwhelms the buffer (which is apparently much smaller than 10GB) and has to fall back on the drive itself (which is relatively slow).

Here's a poorly drawn chart I made showing transfer speeds during a theoretical benchmark run with a drive that has RAPID mode enabled. In this scenario, I'm just guessing that it has a 1GB buffer. The transfer speeds are very high until the buffer runs out in which case the transfer rate falls back down to the normal speed of the drive. Because we have that high burst at the beginning, that average for the test is higher than the drive speed itself. If you were to increase the length of the test even further, you'd probably see results very close to the normal speed of the drive since that short burst from the buffer will be almost completely averaged out in the long run.

upload_2016-5-31_12-32-58.png


Long story short, these high speeds are possible because for that first little bit, you're not going over SATA, you're going over your memory bus. Once SATA takes over, you fall back down to normal speeds.

If you were to transfer a file to your SSD, it would get stored in memory very quickly (and the transfer would be considered "complete"), but in the background, your SSD would slowly write all the data to itself until that buffer was cleared.
Since I like analogies so much, let's say I made two cheeseburgers and gave them for you to eat. You can't eat them both instantly, so begin nibbling on the first cheeseburger while holding the second in your other hand. (Your hands are the memory buffer). From my perspective, I gave you two cheeseburgers and my job is done and I can walk away, but it'll take you time to slowly nibble away on them. Before you finish, I try to give you two more cheeseburgers, but you can't hold anymore, so I have to sit and wait for you to finish those two before I can give you two more. (Full buffer, transfer has to wait).
I think I'm hungry. :laugh:
Seg22 said:
Btw: Sorry for my bad English and the pictures which i made with the phone
Click to expand...
No worries. Welcome to the forums! :toast:
 
Do you think is ok? I mean the HHD is not that old :/
 

Attachments

  • as-ssd-bench TOSHIBA MQ01ABD1 7.19.2016 8-36-07 AM.png
    as-ssd-bench TOSHIBA MQ01ABD1 7.19.2016 8-36-07 AM.png
    29.6 KB · Views: 913
arvamircea said:
Do you think is ok? I mean the HHD is not that old :/
Click to expand...
You're running a benchmark for SSDs on a mechanical drive. Those are normal speeds. :toast:
 
GShark said:
can u please tell me why your samsung 850 evo drive is surpassing his own specs by a lot ?
Click to expand...

Samsung rapid mode look up two posts from your own
 
I'll get there soon bummer cant run in Rapid mode
 

Attachments

  • as-ssd-bench NVMe Samsung SSD 9.27.2016 11-29-56 PM.png
    as-ssd-bench NVMe Samsung SSD 9.27.2016 11-29-56 PM.png
    30.8 KB · Views: 660
My 2x 840 Pro RAID0
 

Attachments

  • as-ssd-bench Intel Raid 0 Vol 11.8.2016 12-59-46 AM.png
    as-ssd-bench Intel Raid 0 Vol 11.8.2016 12-59-46 AM.png
    29.8 KB · Views: 600
  • Like
Reactions: xvi
Samsung 850 EVO 512Gb Pro Rapid Mode .





 

Attachments

  • as-ssd-bench Samsung SSD 850 11.11.2016 12-32-41 PM.png
    as-ssd-bench Samsung SSD 850 11.11.2016 12-32-41 PM.png
    31.5 KB · Views: 677
as-ssd-bench Samsung SSD 850 11.11.2016 12-32-41 PM.png
 
2x500GB 850 Evo's in RAID 0. I wouldn't know if that was a good score or not...

ssdbench.jpg
 
Kingston Predator 240GB
as-ssd-bench Kingston SHPM228 11.11.2016 21-01-32.jpg
 
Last edited:
I don`t no what`s the problem...i have it on de PCIe X16 @X4 :twitch:
 
Incompressible Data Transfer (AS-SSD and CrystalDiskMark)
240GB — 1290MB/s Read and 600MB/s Write
480GB — 1100MB/s Read and 910MB/s Write
960GB — 1300MB/s Read and 1000MB/s Write

These are the specs...
 
Yeah it's the 960GB version doing 1000 write.
 
yes, and mines is the 240GB... its enough for me. I`ve got e normal Gaming/Bench RIG :)
 
Just checking on what people were seeing...

512gb m.2 Samsung 951's in Raid 0 on an MSI GT73VR
 

Attachments

  • asssd.jpg
    asssd.jpg
    86.7 KB · Views: 469
These are not all AS SSD and what are your thoughts with the native windows 10 NVMe driver?
These are Windows vs Samsung and also with crystaldisk and ATTO as also wondered why such wild difference?
I can do the test but maybe someone else can provide more info on the results.

AS SSD Standard NVM Express Controller.PNG AS SSD Samsung NVMe controller.PNG AS SSD Primocache+Samsung driver.PNG


ATTO Standard NVM Express Controller.PNG ATTO Samsung NVMe controller.PNG ATTO Promocache+Samsung driver.PNG


Crystaldisk 5 Standard NVM Express Controller.PNG Crystaldisk 5 Samsung NVMe controller.PNG Crystaldisk 5 Primocache+Samsung driver.PNG

Left to right Windows standard VVM Express controller, Samsung NVMe 2.0, Samsung + Primocache
I was a bit worried about purchasing an OEM Samsung but if you do a search for Samsung NVME WHQL or Samsung NVMe you will find them like here https://onedrive.live.com/?id=5014229B9E752333!30939&cid=5014229B9E752333
just could find them on the samsung site and the ones I did use ignored the OEM version but those work.

anyone know why the benchmarks vary so much?
 
Last edited:
Here's mine, Crucial M550 mSATA SSD in a SYBA SATA-3 enclosure, connected to a 3.5" to 2.5" bay converter with fan blowing on all the time, otherwise would be running at 60-70C, no matter where installed. The folks at Crucial says this is 'normal', whereas my 2.5" version of the same model runs at a cool 32-34C & was $50 less after all said & done.

vmK7r8R.png


As soon as I post here again, the next will blow the roof off, and w/out artificial enhancements/gimmicks such as RAPID to boost my specs. In fact, some of the PC benchmark sites disqualifies those who are using this type of drive caching, one of the more popular, User BenchMark. Sometimes they'll even dock over a HDD that's using Intel RST because it's being cached. Here, while they say while the PC is 'likely operated by a technical master', note that 'excluding HDD' is on red near the top. For crying out loud, it's a 2TB WD Gold with an internal 128MB cache, there's not that many benches of the drive on their site.

http://www.userbenchmark.com/UserRun/3625410

http://www.userbenchmark.com/UserRun/3618970

Cat
 
You must log in or register to reply here.
Back
Top