USB 3.0 Performance

falagund · 03-23-2013, 11:57 AM

Looking at the comments on the article The Ultimate Storage Solution? inspired me to do some research into USB3. I was guessing that most of the speed loss (5Gb/s =<bits to byte>> 625MB/s =<8/10 encoding>> 500MB/s) from 500MB/s was due to the device being slow (flash/or the disk controller) but apperently there is issues with the protocol itself.

This article (i do like tomshardware, he has done many good reviews int he past) explains more problems with the protocol. Apparently the USB3 protocol stats that 400MB/s is realistic after more overhead (flow control, link status, framing,...). Then also the data communication protocol (BOT Bulk-only transport). This doesn't allow for command queuing, (think assembly line). There is a new replacement for BOT but it's not widely implements (including no drivers in windows 7) called UAS.

Faster USB 3.0 Performance: Examining UASP And Turbo Mode : USB 3.0: Dude, Where's My Speed?

Anyways i thought it was a good read.

Adam · 03-23-2013, 12:09 PM

Originally posted by falagund

mments on the article The Ultimate Storage Solution? inspired me to do some research into USB3. I was guessing that most of the speed loss (5Gb/s => 625MB/s =<8/10 encoding>> 500MB/s) from 500MB/s was due to the device being slow (flash/or the disk controller) but apperently there is issues with the protocol itself. This article (i do like tomshardware, he has done many good reviews int he past) explains more problems with the protocol. Apparently the USB3 protocol stats that 400MB/s is realistic after more overhead (flow control, link status, framing,...). Then also the data communication protocol (BOT Bulk-

My rule of thumb is to halve the max theoretical speed and treat that as the best-case speed in practice. And you really need some high-end PC hardware to be able to take advantage of these extreme speeds

falagund · 03-23-2013, 12:50 PM

For networking 100Mb/s i have seen and 11MB download speed (that is as close to the 12.5MB/s theoretical max)
sata2 i have measured speeds of 200MB/s in a PCIexpress x1 (close to 250 theoretical max)
on my 10ks i have measured 107MB/s (close to the 125MB/s spec)

This sort of closeness should be expected from standards.

Anvh · 03-23-2013, 01:03 PM

So far i know, these speeds are the measured speeds up AND down, so it's actually split.

03-23-2013, 11:57 AM	#1
falagund New Member Join Date: Aug 2011 Location: Rhode Island Posts: 9	USB 3.0 Performance Looking at the comments on the article The Ultimate Storage Solution? inspired me to do some research into USB3. I was guessing that most of the speed loss (5Gb/s =<bits to byte>> 625MB/s =<8/10 encoding>> 500MB/s) from 500MB/s was due to the device being slow (flash/or the disk controller) but apperently there is issues with the protocol itself. This article (i do like tomshardware, he has done many good reviews int he past) explains more problems with the protocol. Apparently the USB3 protocol stats that 400MB/s is realistic after more overhead (flow control, link status, framing,...). Then also the data communication protocol (BOT Bulk-only transport). This doesn't allow for command queuing, (think assembly line). There is a new replacement for BOT but it's not widely implements (including no drivers in windows 7) called UAS. Faster USB 3.0 Performance: Examining UASP And Turbo Mode : USB 3.0: Dude, Where's My Speed? Anyways i thought it was a good read.

03-23-2013, 12:09 PM	#2
Adam Administrator Site Webmaster Join Date: Sep 2006 Location: Arizona Photos: Gallery \| Albums Posts: 51,595	Originally posted by falagund mments on the article The Ultimate Storage Solution? inspired me to do some research into USB3. I was guessing that most of the speed loss (5Gb/s => 625MB/s =<8/10 encoding>> 500MB/s) from 500MB/s was due to the device being slow (flash/or the disk controller) but apperently there is issues with the protocol itself. This article (i do like tomshardware, he has done many good reviews int he past) explains more problems with the protocol. Apparently the USB3 protocol stats that 400MB/s is realistic after more overhead (flow control, link status, framing,...). Then also the data communication protocol (BOT Bulk- My rule of thumb is to halve the max theoretical speed and treat that as the best-case speed in practice. And you really need some high-end PC hardware to be able to take advantage of these extreme speeds Adam PentaxForums.com Webmaster (Site Usage Guide \| Site Help \| My Photography) PentaxForums.com server and development costs are user-supported. You can help cover these costs by donating or purchasing one of our Pentax eBooks. Or, buy your photo gear from our affiliates, Adorama, B&H Photo, KEH, or Topaz Labs, and get FREE Marketplace access - click here to see how! Trusted Pentax retailers:

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Pretec USB 3.0 Card reader Speed tests	Chaos_Realm	Digital Processing, Software, and Printing	3	06-11-2012 03:49 PM
USB 2.0 Support from the Pentax site - chuckle	Aknot	Pentax DSLR Discussion	7	03-29-2011 10:49 PM
Breaking my hand-holding records with 0.3, 0.6, 0.8 and 1 sec shutter	hinman	Pentax DSLR Discussion	2	01-02-2008 10:56 PM
USB Download Speed with Remote Assistance 3.0 with K10D v1.30	SBAYERL	Photographic Technique	4	12-29-2007 04:02 AM
Sandisk Extreme USB 2.0 card reader	Cideway	Pentax Camera and Field Accessories	4	06-11-2007 08:27 AM

03-23-2013, 12:50 PM	#3
falagund New Member Join Date: Aug 2011 Location: Rhode Island Posts: 9 Original Poster	For networking 100Mb/s i have seen and 11MB download speed (that is as close to the 12.5MB/s theoretical max) sata2 i have measured speeds of 200MB/s in a PCIexpress x1 (close to 250 theoretical max) on my 10ks i have measured 107MB/s (close to the 125MB/s spec) This sort of closeness should be expected from standards.

03-23-2013, 01:03 PM	#4
Anvh Veteran Member Join Date: Sep 2011 Posts: 4,616	So far i know, these speeds are the measured speeds up AND down, so it's actually split.