Lexar 64GB Professional UHS-II SD Card + M240p

[cnick6]

Advertisement (gone after registration)

I was reading in the Secure Digital specification trying to learn the difference between UHS-I and UHS-II and what makes them tick -- until a thought hit me.  My thought was a simple question -- what part of the SD system is responsible for pushing data?  The memory card or the SD host.  The answer is somewhat obvious -- both -- but it doesn't answer the question of can I get a speed improvement on a UHS-I SD Host device when using a UHS-II SD card.  

 

The answer appears to be -- yes!

 

I purchased a Lexar Professional 64GB UHS-II SD card rated at 300MB/s maximum transfer rate.  I also purchased a Lexar UHS-II reader which works off a USB 3.0 port.  Transferring data between my PC and the card was pretty impressive.  282MB/s was the fastest I saw but the average was closer to 225MB/s.  I was able to transfer a 4GB ISO image file in about 10 seconds.  So yeah, UHS-II is fast.

 

I have no way to accurately measure the time but in looking at the storage LEDs on my two cameras -- there was a measurable (by the naked eye) -- time difference between the UHS-I and UHS-II cards.  (The UHS-I card is a Sandisk Extreme Pro 64GB.)  Both are formatted exFAT using SDFormatter 4.0.

 

Unscientific measurement data:

 

(Simply taking 10 pictures of well-lit scenes at the same exposure settings and watching the write LED...)

 

Sony A7R2 == 1/2 second difference (5s to 5.5s)

Leica M240P == 1 second difference (6s to 7s)

 

PS. I should point out that the Sony is writing JPEG and the Leica is writing DNGc+JPEG.  So ~30MB vs ~40MB of data 

 

I looked at the SD 5.0 specification again (physical data) and it shows that SD cards have a SD controller on them which is responsible for the data flow.  UHS-II cards actually have a controller chip with two sub-controllers (one for SD thru UHS-I speeds and the other for UHS-2 speed.)

 

My theory is that the SD Host is akin to a large water pipe.  During maximum UHS-I speed, the water flowing is about one-half the size the pipe; however, on a UHS-II card, the water flowing is more like 3/4 the size.  Per the specification, the SD host has no traffic cop.  As long as data packets are written perfectly, it'll accept any extra flow.

 

If there are any electrical engineers with PHDs in microelectronic storage mediums, then I'd love to hear your input.  

 

FWIW, I've switched my Leica M240p to a 64GB UHS-II card.  I'll keep an eye out for any issues.

 

 

Cheers,

-Nick

Edited October 26, 2016 by cnick6

[kaiserphoenix]

Interesting, in my extremely subjective tests, i felt that UHSII cards actually ran slower than on a UHSI card....But I think I probably need to test it properly. If your tests are conclusive, it would be a huge positive news!

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[cnick6]

Kaiser, I think this definitely would require more testing and on multiple devices.  I wouldn't be shocked to discover firmware that limits the SD transfer speed.