Jump to content











Photo

Is Transcend Premium 300x UHS-I card overrated?


  • Please log in to reply
55 replies to this topic

#26 Holmes.Sherlock

Holmes.Sherlock

    Gold Member

  • Team Reboot
  • 1444 posts
  • Location:Santa Barbara, California
  •  
    United States

Posted 15 June 2014 - 07:38 AM

BTW, I've added info to post #8 and reformulated it, aiming to make it clearer, please do give it another read.

 
Which card does the below stats refer to?
 
Adapter: Dell U2410F Monitor Card-Reader on USB 2.0
    --------------------------------------------------
    CrystalDiskMark 2.2 (C) 2007-2008 hiyohiyo
          Crystal Dew World : http://crystalmark.info/
    --------------------------------------------------

       Sequential Read :   20.888 MB/s
      Sequential Write :   16.847 MB/s
     Random Read 512KB :   20.503 MB/s
    Random Write 512KB :    9.598 MB/s
       Random Read 4KB :    4.393 MB/s
      Random Write 4KB :    2.073 MB/s

             Test Size : 100 MB <<<


#27 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 15 June 2014 - 07:48 AM

To the same full-card SDHC SANDISK Extreme (45 MB/s, Class 10, UHS-I) 32 GB I used for all my other tests.

I see the mention to the card was truncated when I copy/pasted... that's corrected now, thanks.

 

I used however the card-reader that comes with the monitor, instead of the Transcend RDS-5, just to illustrate the fact that all USB 2.0 adapters do give similar results on USB 2.0, while a USB 3.0 adapter gives much better results, even on USB 2.0, for the same card, unless the card is not good at all.



#28 Holmes.Sherlock

Holmes.Sherlock

    Gold Member

  • Team Reboot
  • 1444 posts
  • Location:Santa Barbara, California
  •  
    United States

Posted 15 June 2014 - 08:00 AM

I used however the card-reader that comes with the monitor, instead of the Transcend RDS-5, just to illustrate the fact that all USB 2.0 adapters do give similar results on USB 2.0,...

 

Is Dell U2410F Monitor Card-Reader an USB 2.0 one?

 

 

 

...while a USB 3.0 adapter gives much better results, even on USB 2.0, for the same card, unless the card is not good at all.

 

Don't have any idea how to explain this. Theoretically an 3.0 reader should default to 2.0 when used with a 2.0 port.



#29 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 15 June 2014 - 08:51 AM

Is Dell U2410F Monitor Card-Reader an USB 2.0 one?

Yes, it is.
 

Don't have any idea how to explain this. Theoretically an 3.0 reader should default to 2.0 when used with a 2.0 port.

Well, USB 2.0 is from 2000, and even at that time it's questionable whether it represented the state-of-art of data transmission, since other matters (viz. being a one type fits all interface, for instance) seem to have been paramount in the minds of the people who defined the standard. But USB 2.0 controllers were made to cope with the USB 2.0 standard, whereas those for USB 3.0 needed to be much faster, internally, to be able to abide by the newer USB 3.0 standard. This also means they're insanely faster than required to deal with USB 2.0 and, when operating in "USB 2.1" mode use the USB 2.0 signaling and protocols, but present virtually zero hysteresis and minimal overhead, because "USB 2.1" do not require them to slow down more than what needed to cope with the 4 cable interface, and what will end up limiting the final transference-ratio is the motherboard soutbridge USB controller or the add-on card (when present). Hence, differently from an USB 2.0 adapter, a USB 3.0 adapter connected to a USB 2.0 port will present almost no overhead at all. It does default to "2.1" mode, but remains faster than any real USB 2.0 would ever be. **The same applies equally to USB 3.0 pendrives in "USB 2.1" mode**



#30 Wonko the Sane

Wonko the Sane

    The Finder

  • Advanced user
  • 16066 posts
  • Location:The Outside of the Asylum (gate is closed)
  •  
    Italy

Posted 15 June 2014 - 02:23 PM

Just to add some uncertainty to the matter :whistling:  how can anyone say that a USB 2.0 is really a USB 2.0 (and not - as an example - a USB 2.37 or a USB 2 1/4 :w00t:)?

I mean, is the USB "standard" a "forced speed cap" or a "top speed" or a "minimal speed"?

 

The USB 2.0 standard (besides having been managed in a far from "standard" way) is not really-really "static" and unchanged since 2000:

http://www.usb.org/d...ocs/usb20_docs/

 

The "gap" between theoretical USB 2.0 speed and USB 3.0 speed is so large that I wouldn't be entirely surprised if - when comparing the performance of a USB 2.0 (adapter/interface/port) actually built in (say) 2001 against a correspondent piece of hardware built in (still say) 2012 - it would come out that the first barely reaches standard speeds while the latter exceeds them.

 

Another thing that we really do not know (or know in detail) is how exactly the "Turbo" drivers for selected USB 2.0 devices manage on same hardware to get a speed increase (and how much of that has been later - hypothetically - transferred to the actual firmware/controller of the device/port):

http://reboot.pro/to...695-turboflash/

 

:duff:

Wonko



#31 Holmes.Sherlock

Holmes.Sherlock

    Gold Member

  • Team Reboot
  • 1444 posts
  • Location:Santa Barbara, California
  •  
    United States

Posted 16 June 2014 - 03:58 AM

Let me summarize once again certain points
  • An USB 3.0 card reader (e.g. Kingston MobileLite G3 (FCR-MLG3)) performs better than an USB 2.0 card reader (e.g. Transcend RDS-5), both running on USB 2.0 (post# 4)
  • The READ speed (e.g. 30.495 MB/s) of an UHS-I card (e.g. SANDISK Extreme (45 MB/s, Class 10, UHS-I) 32 GB) can max-out an USB 3.0 adapter (e.g. Kingston MobileLite G3 (FCR-MLG3)) running on USB 2.0 (practical maximum 35 MB/s) (post# 4)
  • With all other parameters (e.g. card, reader, host) kept intact, sequential write values tend to approach the sequential read values as the test-file size gets bigger (post# 8)


#32 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 16 June 2014 - 08:24 AM

Yes, that's right! But, as far as I can see, what must max-out in 2. is the USB 2.0 controller on the machine side, not the USB 3.0 adapter itself... and probably the WRITE speed can be maxed out, too, with the right experimental set-up.

 

Now, I know, and it's written in many places, including the English Wikipedia, that the transfer-rate practical maximum for USB 2.0 is 35 MB/s, and I admit the maximum I've ever been able to measure was 32.7 MB/s (on my A7V600-X/Athlon XP 3000+ @2337 MHz), but I've never ever seen anyone even try to explain why it's so. The standard talks about 480 Mb/s, which could be 60 MB/s (or 48 MB/s in case a 8-in-10 bit encoding is perhaps used). What is the nature of the overhead on the USB 2.0 interface? I did search the web for any explanation, but came back empty-handed, so far... Then again, since now that we're discussing it, would Wonko kindly give it a try, please?     



#33 Wonko the Sane

Wonko the Sane

    The Finder

  • Advanced user
  • 16066 posts
  • Location:The Outside of the Asylum (gate is closed)
  •  
    Italy

Posted 16 June 2014 - 09:27 AM

dencorso, on 16 Jun 2014 - 10:24 AM, said:
The standard talks about 480 Mb/s, which could be 60 MB/s (or 48 MB/s in case a 8-in-10 bit encoding is perhaps used). What is the nature of the overhead on the USB 2.0 interface? I did search the web for any explanation, but came back empty-handed, so far... Then again, since now that we're discussing it, would Wonko kindly give it a try, please?

You want a theoretical explanation of a practical issue? :w00t:

The data is transferred in "frames", each frame is made of three "packets" (of which the middle one or "data" one) is "optional":
http://www.beyondlog...ml#USBProtocols
http://www.usbmadesi...co.uk/ums_6.htm
http://www.usbmadesi...co.uk/ums_7.htm

From what I understand, the bus is "half-duplex", in an extremely simplified way, think of the differences between a "normal" telephone conversation when compared against one over a walkie-talkie:

Quote
Hi, I am the motherboard controller, anyone connected to this bus? Over.
Hi, I am a USB Mass Storage device, are you a motherboard controller? Over.
Hi I am the motherboard controller, Roger and Over.
Ok, I am going to send you some data, Over.
Fine, I am ready to receive the data, Over.
Here is the data, Over.
Ok, got them, Roger and Over.
You ready for some more data? Over.
Sure, send'em over. Roger and Over.
Here is the data, Over.
Ok, got them, Roger and Over.
You still there or some demented user disconnected us in the meantime? Over.
I am still here, Over.
Good, I am going to send you some more data, Over.
Roger, keep'em coming. Over.
Have you got them? Over
Sure, go on. Over.
I am ready to send some more dats, you still there? Over.
....

The "Over" or "Roger and Over" pieces of conversion (excluding the actual data transmitted) correspond loosely to the SOF (start of frame) packets (and take place every 125 microseconds):
http://www.keil.com/...__protocol.html
See also:
http://uk.farnell.co...dev_mistake.pdf



Quote
As with any electronic system, designers want the best performance possible.
Unfortunately, with USB, many designers begin their design believing they are
going to get the full 1.5, 12, or 480 Mbit/sec performance from their system – this
is a bad assumption. There are several reasons why your device will never be
able to use all of this bandwidth. First of all, the USB bus is shared among
several users. Even if you are plugged into different ports on the motherboard,
you are probably sharing the same host controller as all of the other devices on
the bus, so your device is sharing the USB bus bandwidth with all of the other
devices.

Second, USB is a packetized protocol where longer blocks of data are divided
into 512-byte packets. Each packet contains a header identifying the packet
contents, and a CRC at the end of the packet for data integrity. Each packet also
requires an ACK from the other side of the link. Start of Frame (SOF) packets
are sent every 125 uSec (microframe) to maintain timing on the bus. The net
effect of this is that the theoretical maximum bandwidth of USB is 13 bulk
packets per microframe, or 53,248,000 bytes/second. Even this limit is not
achievable with current host controllers, which can receive 10 bulk
packets/microframe or send 8 bulk packets/microframe.

AFAICT, the mentioned Turboflash (and the Windows 7 "Registry hack") are related to using not the "normal" BULK Transfer protocol but another one that has less overhead for "BIGgish" data chunks):
http://reboot.pro/to...rboflash/page-2

:duff:
Wonko
  • dencorso likes this

#34 Holmes.Sherlock

Holmes.Sherlock

    Gold Member

  • Team Reboot
  • 1444 posts
  • Location:Santa Barbara, California
  •  
    United States

Posted 16 June 2014 - 05:10 PM

I asked Transcend Support for not-so-promising WRITE speed of the card and here's their response
 

Speed may vary due to host hardware, software and usage.[/size]
[/size]
Our test vaule for TS32GUSDU1 with USB3.0 platform is read: 80MB/s, write:25MB/s.[/size]

For more product FAQ, please refer to the following link:[/size]
http://tw.transcend-...t.asp?exid=1202

 
Now I have queried them the following and awaiting response
 

May I know why the card was rated to be a 300X (45 MB/s) one while neither READ nor WRITE is upper limited by that particular value?

Also, during your lab testing, did you use Transcend USB 3.0 card readers (e.g. TS-RDF5 or TS-RDF8) or some other in-house equipment?


Also now I feel an irresistible desire to kick on the butts of those handful of reviewers on flipkart claiming to experience 45 MB/s, what Transcend itself could not.



#35 Wonko the Sane

Wonko the Sane

    The Finder

  • Advanced user
  • 16066 posts
  • Location:The Outside of the Asylum (gate is closed)
  •  
    Italy

Posted 16 June 2014 - 06:39 PM

Do you really think that you can simply call (or write to) support of a large firm and actually get some support on a real technical question?
Usually you have to escalate two or three levels to find someone who actually knows where his/her towel is. :ph34r:
Whether he/she will be willing to share this piece of info with you is of course another matter.

Generally speaking of course (i.e. not necessarily applying to Transcend, mind you)

:duff:
Wonko

#36 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 16 June 2014 - 08:12 PM

Holmes.Sherlock, on 16 Jun 2014 - 2:10 PM, said:

Our test vaule for TS32GUSDU1 with USB3.0 platform is read: 80MB/s, write:25MB/s.

On post #23 I had inferred your card has seq. read: 84 MB/s and 17 MB/s seq write! Those values do compare fairly well with those quoted by the "Transcend Advanced Support Public Liaison Officer" (lol! aka: lowest possible level support). :)

And I had also said that Transcends are not at all great cards, but not quite bad either, although I maybe didn't say it in so many words. So, now, do I think your card is a typical Transcend? No. I do think it's among the best Transcend has to offer!!!
Do I think Transcends are overrated among the tech community (which was the way I understood your thread-opening question)? No, I don't. Now, do I think it was overrated by the user who reviewed it for that particular site you used to buy the card? Perhaps. Or, perhaps the card he received was really better than the one you've got. From the user reviews at amazon.com, one can clearly see Transcend appears to continue using a packaging even when the actual hardware inside has very different specs than those printed in the packaging (and that's precisely what the line "these specs may change without prior notice from the manufacturer", so commonly encountered in packaging descriptions, may mean).

So, yes, as the saying goes, since way back when: caveat emptor!

#37 Holmes.Sherlock

Holmes.Sherlock

    Gold Member

  • Team Reboot
  • 1444 posts
  • Location:Santa Barbara, California
  •  
    United States

Posted 17 June 2014 - 02:04 AM

On post #23 I had inferred your card has seq. read: 84 MB/s and 17 MB/s seq write! Those values do compare fairly well with those quoted by the "Transcend Advanced Support Public Liaison Officer" (lol! aka: lowest possible level support). :)

 

I have updated the original post with the CrystalDiskmark Benchmark performance of the card on USB 3.0



#38 Holmes.Sherlock

Holmes.Sherlock

    Gold Member

  • Team Reboot
  • 1444 posts
  • Location:Santa Barbara, California
  •  
    United States

Posted 17 June 2014 - 03:18 AM

Do you really think that you can simply call (or write to) support of a large firm and actually get some support on a real technical question?
Usually you have to escalate two or three levels to find someone who actually knows where his/her towel is.

 

...quoted by the "Transcend Advanced Support Public Liaison Officer" (lol! aka: lowest possible level support). :)

 
The values were provided by "Application Engineer, Tech Support Department 2", which is Transcend's Tier-II support as the mail chain I posted on the other thread says.[/size]
 

Whether he/she will be willing to share this piece of info with you is of course another matter.

 
Already he has shared *some* information which is not publicly available. Doesn't any country have any "Right-to-Information" act to make them divulge the information, thereby stopping such malpractice?
 

Or, perhaps the card he received was really better than the one you've got.


The possibility of such an event is too slim. Having a single card outperforming other 99 results seems to be far-fetched.
 

From the user reviews at amazon.com, one can clearly see Transcend appears to continue using a packaging even when the actual hardware inside has very different specs than those printed in the packaging (and that's precisely what the line "these specs may change without prior notice from the manufacturer", so commonly encountered in packaging descriptions, may mean).


When an end-user purchases a item, the wrapper is what one can only look at. If the hardware and packaging differ, I am sure there are laws protecting customer rights. Changing the specification for a brand can never mean tricking a customer to buy a lower-performing hardware wrapping it up inside an eye-candy cover.



#39 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 17 June 2014 - 03:56 AM

When an end-user purchases a item, the wrapper is what one can only look at. If the hardware and packaging differ, I am sure there are laws protecting customer rights. Changing the specification for a brand can never mean tricking a customer to buy a lower-performing hardware wrapping it up inside an eye-candy cover.

 

You're right, of course.
 

Or, perhaps the card he received was really better than the one you've got.

 

What I meant here is a true seq-read 45 MB/s; seq-write 45 MB/s (older) instead of a seq-read 90 MB/s; seq-write 15 MB/s (newer)... but notice that one can contend that what matters is seq-read, in which case 90 MB/s is better than 45 MB/s (because unqualified transfer-rates almost always mean seq-read)... so that the older, having higher seq-write transfer-rate may be better *in my opinion* but not on that of the manufacturer, who nonetheless cannot be construed as disingenious, because he/she can contend the product offered is actually better than announced on the cover... YMMV, as you can see (or, as Pontius Pilatus is said to have put it: what is truth?)...

And just to put things in perspective, along 2014, the state-of-art SDHC UHS-I card from SanDisk changed twice, at least, and they advertised it quite clearly, as you can see in the attached pic (the one I used in my tests is identical looking to the one at extreme left, in the pic):

 

razoredge.png



#40 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 17 June 2014 - 07:25 AM

You want a theoretical explanation of a practical issue? :w00t:


Yes. You put me on the right direction. Your penultimate link, in fact, hit jackpot. :worship: It actually gave the answer, but I failed to do the math at the moment I read it, but all the relevant info was there (and you've even provided a quote of the relevant part)... Afterwards, I found this analysis, which led me even nearer: it concludes that 53.248 MB/s is maximum theoretical number for bulk upstream devices. This number comes from the standard. But they go on to show it's calculated as 13 packets x 512 bytes/packet x 8000 packets/s. Then, one of the comments to that blog post merits quoting:

While there may be space for 13 packets in the microframe, many host controllers cannot actually do that. Way back in 2006, most were limited to 8 out and 10 in--I have no idea if that has changed much since then or not.


All the above info was thus provided twice, in the two above mentioned sources. But it still remained necessary to give one last step and recalculate the theoretical number for an average of 9 (instead of 13) packets per microframe: by doing it, one gets 9 x 512 x 800 = 36.864 MB/s or 35.156 MiB/s... :)

#41 Wonko the Sane

Wonko the Sane

    The Finder

  • Advanced user
  • 16066 posts
  • Location:The Outside of the Asylum (gate is closed)
  •  
    Italy

Posted 17 June 2014 - 10:53 AM


All the above info was thus provided twice, in the two above mentioned sources. But it still reained necessary to give one last step and recalculatethe theoretical number for an average of 9 (instead of 13) packets per microframe: by doing it, one gets 9 x 512 x 800 = 36.864 MB/s or 35.156 MiB/s... :)

Yep :), and this is exactly what I meant here:

Just to add some uncertainty to the matter :whistling: how can anyone say that a USB 2.0 is really a USB 2.0 (and not - as an example - a USB 2.37 or a USB 2 1/4 :w00t:)?
I mean, is the USB "standard" a "forced speed cap" or a "top speed" or a "minimal speed"?

The USB 2.0 standard (besides having been managed in a far from "standard" way) is not really-really "static" and unchanged since 2000:
http://www.usb.org/d...ocs/usb20_docs/

The "gap" between theoretical USB 2.0 speed and USB 3.0 speed is so large that I wouldn't be entirely surprised if - when comparing the performance of a USB 2.0 (adapter/interface/port) actually built in (say) 2001 against a correspondent piece of hardware built in (still say) 2012 - it would come out that the first barely reaches standard speeds while the latter exceeds them.


The protocol allows a MAXimum theoretical number of 13 packets per second.
But this depends of the "computing capabilities" of the hardware.
It would be likely that older hardware is capable of 8 or 9 packets (actual USB 2.0 hardware released in 2001 or so), that in (say) 2006 it got to 10 or 11 packets (USB 2.1/4 ;)) and that now, on "good" brands, is 12 or maybe 13 (USB 2.37 :whistling:)...

:duff:
Wonko

#42 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 18 June 2014 - 11:07 AM

Here's a little more food for thought...
===================================================
OCZ ATV Turbo 8 GB USB 2.0 (SMI SM325 - 4x 2GB SLC)
Intel Z68 onchip USB 2.0 Controller 
i7 3770k / Asus P8Z68-V LX - Win XP SP3
--------------------------------------------------
CrystalDiskMark 2.2 (C) 2007-2008 hiyohiyo
      Crystal Dew World : http://crystalmark.info/
--------------------------------------------------

    Test File Size :      100 MB       1000 MB
   Sequential Read :   29.425 MB/s   29.428 MB/s
  Sequential Write :   21.680 MB/s   21.704 MB/s
 Random Read 512KB :   29.304 MB/s   29.386 MB/s
Random Write 512KB :    9.744 MB/s    9.576 MB/s
   Random Read 4KB :    5.195 MB/s    5.523 MB/s
  Random Write 4KB :    0.156 MB/s    0.146 MB/s


===================================================
OCZ ATV Turbo 8 GB USB 2.0 (SMI SM325 - 4x 2GB SLC)
Onboard ASMedia 1042 Controller USB 3.0 
i7 3770k / Asus P8Z68-V LX - Win XP SP3
--------------------------------------------------
CrystalDiskMark 2.2 (C) 2007-2008 hiyohiyo
      Crystal Dew World : http://crystalmark.info/
--------------------------------------------------

    Test File Size :      100 MB       1000 MB
   Sequential Read :   35.951 MB/s   35.535 MB/s
  Sequential Write :   24.975 MB/s   24.961 MB/s
 Random Read 512KB :   35.608 MB/s   35.357 MB/s
Random Write 512KB :   10.066 MB/s   10.194 MB/s
   Random Read 4KB :    9.699 MB/s   13.615 MB/s
  Random Write 4KB :    0.167 MB/s    0.144 MB/s


===================================================
OCZ ATV Turbo 8 GB USB 2.0 (SMI SM325 - 4x 2GB SLC)
STW R001 PCI-e add-on USB 3.0 Card (Renesas uPD720201)
i7 3770k / Asus P8Z68-V LX - Win XP SP3
--------------------------------------------------
CrystalDiskMark 2.2 (C) 2007-2008 hiyohiyo
      Crystal Dew World : http://crystalmark.info/
--------------------------------------------------

    Test File Size :      100 MB       1000 MB
   Sequential Read :   35.869 MB/s   35.857 MB/s
  Sequential Write :   23.987 MB/s   24.022 MB/s
 Random Read 512KB :   35.591 MB/s   35.782 MB/s
Random Write 512KB :   10.029 MB/s   10.152 MB/s
   Random Read 4KB :    9.495 MB/s   13.981 MB/s
  Random Write 4KB :    0.165 MB/s    0.146 MB/s


===================================================
OCZ ATV Turbo 8 GB USB 2.0 (SMI SM325 - 4x 2GB SLC)
StarTech PCIUSB3S2 PCI add-on USB 3.0 Card (NEC uPD720200) 
Athlon XP 3000+ @ 2337 MHz / Asus A7V600-X - Win XP SP3
---------------------------------------------------
CrystalDiskMark 2.2 (C) 2007-2008 hiyohiyo
      Crystal Dew World : http://crystalmark.info/
---------------------------------------------------

    Test File Size :      100 MB         1000 MB
   Sequential Read :   33.522 MB/s     33.894 MB/s
  Sequential Write :   20.363 MB/s     20.514 MB/s
 Random Read 512KB :   33.040 MB/s     34.017 MB/s
Random Write 512KB :    9.126 MB/s      9.703 MB/s
   Random Read 4KB :    7.801 MB/s     12.094 MB/s
  Random Write 4KB :    0.155 MB/s      0.145 MB/s

         
===================================================
OCZ ATV Turbo 8 GB USB 2.0 (SMI SM325 - 4x 2GB SLC)
VIA 8237 onchip USB 2.0 Controller 
Athlon XP 3000+ @ 2337 MHz / Asus A7V600-X - Win XP SP3
---------------------------------------------------
CrystalDiskMark 2.2 (C) 2007-2008 hiyohiyo
      Crystal Dew World : http://crystalmark.info/
---------------------------------------------------

    Test File Size :      100 MB         1000 MB
   Sequential Read :   32.630 MB/s     32.614 MB/s
  Sequential Write :   26.449 MB/s     26.980 MB/s
 Random Read 512KB :   32.709 MB/s     32.677 MB/s
Random Write 512KB :   10.739 MB/s     11.366 MB/s
   Random Read 4KB :    8.136 MB/s      8.030 MB/s
  Random Write 4KB :    0.161 MB/s      0.144 MB/s


 
Please do bear in mind that the Asus A7V600-X is a motherboard from 2003 and the Asus P8Z68-V LX is from 2011. Since my objective was to test USB 2.0 (or USB 2.0 compatibility mode for the USB 3.0 systems) I used for all the tests an OCZ ATV Turbo 8GB USB 2.0 pendrive, which is the fastest USB 2.0 pendrive I've ever tested and possibly the fastest 8GB USB 2.0 pendrive ever produced (and already out-of-market for quite some years). So, one thing seems sure: the USB 2.0 controllers didn't improve any from 2003 to 2011... in fact, they may even have become somewhat worse, but it's difficult to assert that because the machines are very different, although it's possible to conclude their performance is just about the same.

#43 Holmes.Sherlock

Holmes.Sherlock

    Gold Member

  • Team Reboot
  • 1444 posts
  • Location:Santa Barbara, California
  •  
    United States

Posted 18 June 2014 - 12:47 PM

Do you really think that you can simply call (or write to) support of a large firm and actually get some support on a real technical question?
Usually you have to escalate two or three levels to find someone who actually knows where his/her towel is. :ph34r:
Whether he/she will be willing to share this piece of info with you is of course another matter.

 

For our UHS-I products of microSD card, the rate show our max transfer data speed
Ultimate series have more faster write speed than premium series, so we distinguish two parts as premium(300X, 45MB/s) and ultimate(600X, 90MB/s).
 
The perfomance is tested by RDF8 card reader and test device as below:
*Processor: Intel Core i5-3570
*OS: Windows 7
*Adapter: TS-RDF8
*USB Host: USB 3.0
*Software: CrystalDiskMark 2.2
*RAM: 4GB



#44 Wonko the Sane

Wonko the Sane

    The Finder

  • Advanced user
  • 16066 posts
  • Location:The Outside of the Asylum (gate is closed)
  •  
    Italy

Posted 18 June 2014 - 02:45 PM

@dencorso

Sorry, but I don't understand the way you draw conclusions from the two tests you made.

If I get this right, let's use for simplicity only sequential read, you have:
  • 29 Mb/s - One random newish motherboard with onboard Intel Z68 USB 2.0 controller
  • 36 Mb/s - Same random newish motherboard with onboard ASMedia 1042 Controller USB 3.0
  • 36 Mb/s - Same random newish motherboard with STW R001 PCI-e add-on USB 3.0 Card (Renesas uPD720201)
  • 33 Mb - One random oldish motherboard with add-on StarTech PCIUSB3S2 PCI USB 3.0 Card (NEC uPD720200)
  • 32 Mb/s -Same random oldish motherboard with onboard VIA 8237 onchip USB 2.0 Controller
All the conclusions that I can draw from these are:
  • USB 3.0 are faster than USB 2.0 even when used with USB 2.0 devices :unsure:
  • The "new" Intel Z68 (or its implementation/drivers/etc. on the specific motherboard) is sensibly slower than an oldish VIA 8237 (or its implementation/drivers/etc. on the specific motherboard)
Point #2 might be something connected to the "specific" chipset/whatever or even something connected to the availability of USB 3.0 (i.e. maybe "slower" USB 2.0 chipsets are cheaper than "faster" ones and since the motherboard has also USB 3.0 connectivity, this is not anymore a priority).

One would need to compare a set of motherboards from the early days of USB 2.0 against a set of "latest before USB 3.0" one to (hopefully) gather some relevant data.

I believe that you need *some device* capable of largely exceeding 35 Mb/s transfer (that would probably be a hard disk - SATA or fastish PATA - connected to a "last generation" USB 2.0 SATA or PATA bridge) AND test it on several different motherboards to build a database from which you can extract data to support the theory.

@Holmes Sherlock
Sure and what actual information (once set aside truisms) does your quote contain?
I will take note that:

Quote
Ultimate series have more faster write speed than premium series...

for future use, however. ;)


:duff:
Wonko

#45 Holmes.Sherlock

Holmes.Sherlock

    Gold Member

  • Team Reboot
  • 1444 posts
  • Location:Santa Barbara, California
  •  
    United States

Posted 18 June 2014 - 02:45 PM

Does Asus P8Z68-V LX have both uSb 2.0 and 3.0 ports?
 

===================================================
OCZ ATV Turbo 8 GB USB 2.0 (SMI SM325 - 4x 2GB SLC)
StarTech PCIUSB3S2 PCI add-on USB 3.0 Card (NEC uPD720200) 
Athlon XP 3000+ @ 2337 MHz / Asus A7V600-X - Win XP SP3
---------------------------------------------------
CrystalDiskMark 2.2 (C) 2007-2008 hiyohiyo
      Crystal Dew World : http://crystalmark.info/
---------------------------------------------------

    Test File Size :      100 MB         1000 MB
   Sequential Read :   33.522 MB/s     33.894 MB/s
  Sequential Write :   20.363 MB/s     20.514 MB/s
 Random Read 512KB :   33.040 MB/s     34.017 MB/s
Random Write 512KB :    9.126 MB/s      9.703 MB/s
   Random Read 4KB :    7.801 MB/s     12.094 MB/s
  Random Write 4KB :    0.155 MB/s      0.145 MB/s

         
===================================================
OCZ ATV Turbo 8 GB USB 2.0 (SMI SM325 - 4x 2GB SLC)
VIA 8237 onchip USB 2.0 Controller 
Athlon XP 3000+ @ 2337 MHz / Asus A7V600-X - Win XP SP3
---------------------------------------------------
CrystalDiskMark 2.2 (C) 2007-2008 hiyohiyo
      Crystal Dew World : http://crystalmark.info/
---------------------------------------------------

    Test File Size :      100 MB         1000 MB
   Sequential Read :   32.630 MB/s     32.614 MB/s
  Sequential Write :   26.449 MB/s     26.980 MB/s
 Random Read 512KB :   32.709 MB/s     32.677 MB/s
Random Write 512KB :   10.739 MB/s     11.366 MB/s
   Random Read 4KB :    8.136 MB/s      8.030 MB/s
  Random Write 4KB :    0.161 MB/s      0.144 MB/s

 

Another point worth noticing is that in case of Asus A7V600-X motherboard, VIA 8237 onchip USB 2.0 Controller provides almost same READ speed but significantly greater WRITE speed compared to StarTech PCIUSB3S2 PCI add-on USB 3.0 Card (NEC uPD720200). Does USB/PCI protocol distinguish between READ and WRITE requests?



#46 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 18 June 2014 - 04:54 PM

The Asus P8Z68-V LX does have both USB 2.0 and 3.0 ports, alright. However, the USB 2.0 are provided onchip by the Z68 southbrigde, which is *precisely* the latest non USB 3.0 top-of-line southbrige created by intel. Asus added an ASMedia 1042 USB 3.0 controller to their motherboard, thus creating one of the 1st USB 3.0 motherboards to hit the market (and it's great for Win XP users, because it does have drivers for Win XP)! However, after I had the system up and running I decided to add 4 more USB 3.0 connectors and added the Renesas uPD720201 PCI-e board, because of my previous good experience with its predecessor, the NEC uPD720200. On the other hand, the VIA VT8237 USB 2.0 is implemented as an extended VT6202 controler, which was arguably the first really reliable USB 2.0 controller to ever hit the market, so my choices were very far from random, being, in fact, very representative of the extremes we were discussing, and hence allowing the inference that the newer controller is, at the very least, not better than the older one. As for the second conclusion, you're quite right, USB 3.0 controllers in "USB 2.1 mode" are really always faster then native USB 2.0 controllers. And, as my pendrive tests already have shown, fast USB 3.0 pendrives conected to USB 2.0 controllers are always faster than the fastest native USB 2.0 pendrives. These are facts.

#47 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 21 June 2014 - 05:25 AM

I was thinking about buying a microSDHC card for some time, but always kept postponing it. This thread, however, provided me the incentive to finally actually go ahead and purchase one. I bought the SanDisk Extreme PLUS microSDHC UHS-I 32 GB Card, which is rated in the frontside of the package simply as "Speed up to 80 MB/s - 533X" and comes with a (passive) micro-to-full-card adapter. Interestingly enough, on the backside of the package, SanDisk decided to be more specific and added a more detailed rating, which reads: "READ up to 80 MB/s - 533X; WRITE up to 50 MB/s - 333X"... Now, my tests show it lives up quite well to the promise, and, in fact, actually slightly exceeds what is promised (I've used the best test conditions as defined by my previous tests with the OCZ ATV Turbo pendrive):

===================================================
SanDisk Extreme PLUS microSDHC UHS-I 32 GB Card
Adapter: Kingston MobileLite G3 (FCR-MLG3) on USB 2.0
Controller: VIA 8237 onchip USB 2.0 
Athlon XP 3000+ @ 2337 MHz / Asus A7V600-X - Win XP SP3
--------------------------------------------------
CrystalDiskMark 2.2 (C) 2007-2008 hiyohiyo
      Crystal Dew World : http://crystalmark.info/
--------------------------------------------------

    Test File Size :      100 MB
   Sequential Read :   27.321 MB/s
  Sequential Write :   26.519 MB/s
 Random Read 512KB :   26.363 MB/s
Random Write 512KB :   19.266 MB/s
   Random Read 4KB :    4.561 MB/s
  Random Write 4KB :    2.063 MB/s

===================================================
SanDisk Extreme PLUS microSDHC UHS-I 32 GB Card
Adapter: Kingston MobileLite G3 (FCR-MLG3) on USB 3.0
Controller: STW R001 PCI-e add-on USB 3.0 Card (Renesas uPD720201)
i7 3770k / Asus P8Z68-V LX - Win XP SP3
--------------------------------------------------
CrystalDiskMark 2.2 (C) 2007-2008 hiyohiyo
      Crystal Dew World : http://crystalmark.info/
--------------------------------------------------

    Test File Size :      100 MB
   Sequential Read :   87.369 MB/s
  Sequential Write :   60.638 MB/s
 Random Read 512KB :   75.722 MB/s
Random Write 512KB :   25.684 MB/s
   Random Read 4KB :    5.896 MB/s
  Random Write 4KB :    2.174 MB/s

===================================================

 

81JHeGmym7L._SY450_.jpg



#48 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 15 December 2014 - 03:52 AM

Here is another, newer, full-size SanDisk SDHC card which lives up quite well to what the manufacturer promises: It's a SanDisk Extreme PRO SDHC UHS-I 32 GB Card, which is rated in the frontside of the package simply as "Speed up to 95 MB/s - 633X", Class 10 and UHS Speed Class 3 (meaning guaranteed minimum sustained sequential write speed of 30 MB/s). My test results are below (as before, I've used the best test conditions as defined by my previous tests with the OCZ ATV Turbo pendrive):
 

===================================================
SanDisk Extreme PRO  SDHC UHS-I 32 GB Card (UHS Speed Class 3)
Adapter: Kingston MobileLite G3 (FCR-MLG3) on USB 2.0
Controller: VIA 8237 onchip USB 2.0 
Athlon XP 3000+ @ 2337 MHz / Asus A7V600-X - Win XP SP3
--------------------------------------------------
CrystalDiskMark 2.2 (C) 2007-2008 hiyohiyo
      Crystal Dew World : http://crystalmark.info/
--------------------------------------------------

    Test File Size :      100 MB
   Sequential Read :   30.701 MB/s
  Sequential Write :   26.931 MB/s
 Random Read 512KB :   29.445 MB/s
Random Write 512KB :   18.042 MB/s
   Random Read 4KB :    4.756 MB/s
  Random Write 4KB :    2.259 MB/s

===================================================
SanDisk Extreme PRO  SDHC UHS-I 32 GB Card (UHS Speed Class 3)
Adapter: Kingston MobileLite G3 (FCR-MLG3) on USB 3.0
Controller: STW R001 PCI-e add-on USB 3.0 Card (Renesas uPD720201)
i7 3770k / Asus P8Z68-V LX - Win XP SP3
--------------------------------------------------
CrystalDiskMark 2.2 (C) 2007-2008 hiyohiyo
      Crystal Dew World : http://crystalmark.info/
--------------------------------------------------

    Test File Size :      100 MB
   Sequential Read :   90.200 MB/s
  Sequential Write :   78.194 MB/s
 Random Read 512KB :   72.008 MB/s
Random Write 512KB :   31.489 MB/s
   Random Read 4KB :    5.849 MB/s
  Random Write 4KB :    2.278 MB/s

===================================================

 

Just like the microSDHC card in my previous post, this Full-Size SDHC card is a great option for running an OS from, their good 4K Random Write Speeds (> 2 MB/s) ensure they will present no speed bottleneck for good OS functioning, and allow for putting the page file on the card confortably, too (of course, karyonix's great DiskMod is required for that).

Attached Files



#49 Brito

Brito

    Platinum Member

  • .script developer
  • 10616 posts
  • Location:boot.wim
  • Interests:I'm just a quiet simple person with a very quiet simple life living one day at a time..
  •  
    European Union

Posted 15 December 2014 - 03:30 PM

this Full-Size SDHC card is a great option for running an OS

 

This is an interesting idea. On my laptop there is still an SD card reader and I like the idea that I can just take the SD card with me to either another machine or just for safety.

 

Also like the fact that imaging 32Gb is fairly simple and SD cards got very affordable. So it would ease the task of having a backup image of the running operating system or to just switch completely as needed.

 

I just wonder if running the OS from an SD card would drain the battery of a laptop faster or if it would actually make the battery last longer since the HDD will not be so active. What do you think?

 

:huh:



#50 Wonko the Sane

Wonko the Sane

    The Finder

  • Advanced user
  • 16066 posts
  • Location:The Outside of the Asylum (gate is closed)
  •  
    Italy

Posted 15 December 2014 - 04:30 PM

The second you said, a hard disk will always drain more power that a solid state device.

 

:duff:

Wonko






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users