[SOLVED] Confused about disk speed between internal and external ssd

[vjovanov]

Hello,

I have 2 computers:

• 1 macbook pro 2015 (with an internal ssd soldered in)
• 1 custom desktop computer on windows 10 (with a KINGSTON SV300S37A240G ssd)

I am used to using an 850 EVO SSD with an adapter as an external hard drive for my macbook pro.
Today I did some speed tests and I was really surprised.

For my internal SSD to my desktop computer connected in SATA III, here are the results (With CrystalDiskMark) : Benchmark image - internal ssd

For my external SSD to my macbook pro connected in USB 3.0, here are the results (With BlackMagic Disk Speed Test) : Benchmark image - external ssd

What surprises me is that from what I understand from the results, from a disk speed point of view, it is more interesting for me to use my macbook pro with an external hard drive than to use my desktop computer which has an internal SSD directly connected to my motherboard.

I thought it was because of a model difference in hard drives. But even so, I'm surprised that my USB3.0 hard drive is faster than my other SATA III hard drive.

Are the results given to me correct? And how can I explain this?

Thank you,
vjovanov

 

[DSzymborski]

First step: run the same benchmark. It defeats the purpose of a benchmark if you mix and match.

 

[Karadjgne]

You can't compare a Kingston drive to a 850 evo. They use different nand and controllers. The Kingston also slows a little somewhere after 50ish% full, so that affects results too.

It's Mustang vs Camero, which is faster. In top end? Off the line? Automatic vs manual shift. Who is the driver, grandma who can't shift faster or pro drag driver?

There's far too many variables to make any sense of why one drive looks to be faster than another. The only thing that really matters is both are a good amount faster than a hdd.

 

[Dennis69]

The technology behind how the drive writes data does not really differ, but the difference is in how it connects to the computer. Because SSDs are much faster than hard drives, your choice of connection is important. Internal SSDs will connect via either SATA or m.2. m.2 is the faster of the two and should be able to make full use of the SSD speed. External SSDs will connect via either USB 2.0, USB 3.0, or eSATA. eSATA and USB 3.0 will give you good speeds, though not as fast as m.2. USB 2.0 will top out around 45 MB/s — not bad, but will only use a fraction of the SSDs performance.
GMGlobalConnect VSP

 

[Karadjgne]

M.2 isn't a format. It's an interface. Basically it's the plug. Sata and NVMe are both able to make use of the plug, with caveats, but that doesn't make them faster. M.2 is Sata. It's no different to the Sata outlets used by a 2.5" drive.

NVMe is faster as that uses pcie, either through the PCH or directly from the cpu.

What slows down external drives is the interface, mainly USB bandwidth limitations, conversions from Sata to USB and back, the controller cards etc.

 

[geofelt]

You are comparing apples to oranges.
Two systems, two different devices, two different apps.

Synthetic tests are nice, but not necessarily applicable to what you do.
What you really should care about is how well your work is done.

See how long it takes do do the same job on each system.
The result may be determined by either the device hardware, or by the cpu running the job.

 

[hotaru.hino]

Running benchmarks on the system drive (in this case C:\) is not recommended because the drive still needs to be used for random things. This will influence the benchmark in a negative way.