• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Is USB 2.0 fast enough?

T

trident9440

n00b
Joined
Oct 23, 2003
Messages
10
I am thinking of buying a USB 2.0 external hard drive, question is will USB2.0 will become the performance bottle neck or it is fast enough for nowadays hard drives?
 
well depends what your using it for. Also, why not get a normal HDD and put it in an enclosure? leaves more options open. But overall using USB2 for a storage HDD is fine. Not a good idea to run progs and games off it, cause it will present a bottleneck if your doin that. But for storing movies and songs it would be fine.
 
I have a similar question.

I have an older 7200 RPM, 25 GB in a Bytec USB 2.0 enclosure that I am sending to my sister. She was originally going to store back-ups of her digital photos and other items on it, but asked me this morning if she could run Quicken from it (she is paranoid about someone hacking into her computer and stealing her financial info.).

Based on trident's statement, would the USB 2.0 connection transfer information fast enough from the HD to be viable?

TIA
 
I run my external enclosure through SATA. That has worked out great for me, cause testing it there is no speed diffrence between running the hdd in the computer and outside in the enclosure. If this is an option for you, definately go with it.
 
setscrew said:
I have a similar question.

I have an older 7200 RPM, 25 GB in a Bytec USB 2.0 enclosure that I am sending to my sister. She was originally going to store back-ups of her digital photos and other items on it, but asked me this morning if she could run Quicken from it (she is paranoid about someone hacking into her computer and stealing her financial info.).

Based on trident's statement, would the USB 2.0 connection transfer information fast enough from the HD to be viable?

TIA
Click to expand...


she will need to install and run her Quicken onto her system hdd and she can setup and store her data files on the removable disk. It should work fine. the only time this bottleneck will effect you is on load time. if you are doing something like games or large software suite those load from the hdd several times in just a few minutes. using onboard IDE or SATA, you will never notice lag. the problem with usb is not only is it's bandwidth lower, the system has to do a bit more processing to get 'er done. so... in teh end, if your just using it for data storage, you should be fine. look at it like a large pendrive. i use a laptop hdd enclosure and i watch my movies off of it on my usb 1.0 laptop and they look fine.
 
USB 2.0 is fast... I dont see any problems using for hard drives...
 
In my case, I am going to install Oracle on my lappy system hdd, but leave the oracle data file on the external hdd. If the speed is 20% lower than local hdd, then that's fine for me. Also wondering if there is any SATA cards can be used by my DELL 600M lappy. It seems the laptop does not have a SATA connector.
Thx.
 
Zardoz said:
USB 2.0 is fast... I dont see any problems using for hard drives...
Click to expand...
Actually, I do see a problem there. You see, USB 2.0's bandwidth is shared between multiple devices on the same bus - and the USB controller hub itself steals part of that bandwidth. And that's not to mention the USB printers, USB scanners and USB mice that are all typically connected at all times to that same USB 2.0 bus. And some of those devices hog so much bandwidth that they practically reduce the available bandwidth of that USB 2.0 bus to no faster than that of a USB 1.1 bus, which is slower than even PIO Mode 0 on an IDE controller! Thus, USB of any currently available speed is not recommended for hard drives that are to be used for running programs. With such device-sharing too commonplace, USB would have to have a theoretical transfer rate in the hundreds of Gb/s (gigabits per second) - about 1000 times faster than USB 2.0's maximum theoretical throughput - in order for a USB hard drive to be competitive in performance to even an internal Ultra ATA/66 hard drive.
 
ok WTH did all your brains go, to the punch bowl?

even the newest drives can't have transfers offplatter over about 40MB/second.

pATA tops at 133MB/sec
sATA tops at 150MB/sec
usb2.0 tops at 400 or 480MB/sec cant remember which.
motherboards have 1 controller per 2 ports, making, guess what, aggregate bandwidth quite high.

OOOOOOOOOOOOH but guess what? PCI bandwidth tops at 132MB/sec and that's what *EVERYTHING* has to go through!!!!

ok so where the hell is the bottleneck exactly? :rolleyes:

by the way im very happy and bubbly im just being a smart @$$ to be funny. sorry dont mean to insult anyone sorry if I did im just havin fun today
merry christmas!!!!!!!!!!!!!
 
480 megaBITs per second, noob. That's 480 divided by 8 bits per byte = 60 megabytes per second maximum THEORETICAL bandwidth. With other devices to share bandwidth with, and lots of CPU and command-level overhead, the actual usable bandwidth is much lower, and the read service time (an important performance metric) is atrocious.

And yes, I've been to the punch bowl a few times already :p
 
LadyJaqie said:
ok WTH did all your brains go, to the punch bowl?

even the newest drives can't have transfers offplatter over about 40MB/second.

pATA tops at 133MB/sec
sATA tops at 150MB/sec
usb2.0 tops at 400 or 480MB/sec cant remember which.
motherboards have 1 controller per 2 ports, making, guess what, aggregate bandwidth quite high.

OOOOOOOOOOOOH but guess what? PCI bandwidth tops at 132MB/sec and that's what *EVERYTHING* has to go through!!!!

ok so where the hell is the bottleneck exactly? :rolleyes:
Click to expand...
Okay, so you're comparing apples and pears, in this case. You see, SATA supports only one device per channel, and the two-device-per-channel P-ATA has supported independent device timing with compatible drives since the advent of ATA/33. However, USB does not support independent device timing at all whatsoever - in fact, each USB device must steal a portion of the maximum available bandwidth. And some older USB devices steal so much that the entire USB 2.0 bus drops to USB 1.1 speed.

And the maximum bandwidth of USB 2.0 is NOT 480MB/sec, but 480Mb/sec (note that there's a BIG difference between MB/sec and Mb/sec, where 1Mb/sec equals 0.125MB/sec, or 1MB/sec equals 8Mb/sec). Thus, the maximum theoretical bandwidth of USB 2.0 is actually only 60MB/sec total (calculated from 480Mb/sec) - and that same 60MB/sec bandwidth is shared between all of the USB devices connected to the same bus, including the USB controller itself.

As for your statement that "motherboards have 1 controller per 2 ports, making, guess what, aggregate bandwidth quite high," that depends on the motherboard. Most motherboards have 1 controller that's shared between all of the USB ports on that motherboard combined, whether that motherboard has 2 USB ports or 8 USB ports.

On the other hand, if USB 2.0 delivers "480 MB/sec," as you said, then P-ATA would actually deliver "1.06 GB/sec," and SATA would deliver "1.2GB/sec."
 
hehe I got owned! sorry guys. Im just all happy go lucky and as I said I was jsut being silly.
but still for hard drives its not so bad of a bus...for a single drive.
 
LadyJaqie said:
hehe I got owned! sorry guys. Im just all happy go lucky and as I said I was jsut being silly.
but still for hard drives its not so bad of a bus...for a single drive.
Click to expand...
Okay, for a hard drive that's being used strictly as data storage... ;)

By the way, the sequential physical read transfer rate of the newer external hard drives - which are almost always 7200rpm hard drives - actually far exceed the maximum sustained bandwidth of USB 2.0, even if that external hard drive is the only device connected to that USB 2.0 port.
 
E4g1e said:
Okay, for a hard drive that's being used strictly as data storage... ;)

By the way, the sequential physical read transfer rate of the newer external hard drives - which are almost always 7200rpm hard drives - actually far exceed the maximum sustained bandwidth of USB 2.0, even if that external hard drive is the only device connected to that USB 2.0 port.
Click to expand...
Yes, but 60MB/s is probably enough. All that's likely to be on the bus, unless I miss my guess, is the disk. KB/mouse are usually on a 1.1 bus, which is believe is seperate entirely.

In response to great-great-grandparent, yes, it's fast enough. Don't worry about it . :D
 
unhappy_mage said:
Yes, but 60MB/s is probably enough. All that's likely to be on the bus, unless I miss my guess, is the disk. KB/mouse are usually on a 1.1 bus, which is believe is seperate entirely.

In response to great-great-grandparent, yes, it's fast enough. Don't worry about it . :D
Click to expand...
I've found that newer systems actually use the very same (commonly shared) controller for both USB 2.0 and USB 1.1 operation - and USB 1.1 devices actually steal bandwidth from the USB 2.0 bus on such newer systems. The systems in which you've had experience with used separate dedicated ports for USB 1.1 and USB 2.0 and which used separate controllers for different ports (a good example of the separate-USB-bus setup is the two-controller configuration on the Asus A7V333, whose VIA KT333 chipset's native USB controller supported only USB 1.1 and then added USB 2.0 support via a separate chipset, also from VIA). Those setups like yours are extremely rare nowadays.
 
You must log in or register to reply here.
Back
Top