• 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.

Physicists Set Record for Network Data Transfer

aronesz said:
Actually, the math is incorrect. =P

186 Gbit/s divided by 8 bits per byte = 23.25 GBytes/s ;)
Click to expand...

The math is indeed correct and I had written out an explanation before why I did it that way and decided to drop it, but seeing as you want to keep correcting me. Here is your explanation:

Again, it fully depends on your methodology. Anytime I see bytes being used, it is usually with binary prefixes (multiples of 1024) instead of metric (multiples of 1000), whereas when bits are used it is usually with straight metric system calculations. I see this even when dealing with communications which is weird. In fact even in my advanced networking technologies courses the professors would accept either method as long as you explained how you did it. Even though the binary prefix is MiB or GiB, it is commonly written as MB or GB still, even in textbooks.

So if you have a metric calculation with bits that is 1Gb, that is 186,000,000,000 bits. Whereas when I see it specified in GB it usually applies the prefix meaning 1,073,741,824 bytes. There are tons of calculators out there that do this such as:

http://www.speedguide.net/conversion.php

and if you put in 186000000000 bits, it comes out as 186 Gb and 21.6 GB.
 
I forgot to mention that metric is used for communications where they tend to stay with bits, whereas binary prefix is used with storage. But seing as transfer speeds may apply to storage or communications, they tend to stay with bits for communication and bytes with storage transfers. So if you are making a measurement of storage transfer that is where the calculation comes in. And when referencing confusion as to why communication companies do not use storage transfer speeds, this should help clarify some of that.
 
NoOther said:
The math is indeed correct and I had written out an explanation before why I did it that way and decided to drop it, but seeing as you want to keep correcting me. Here is your explanation:

Again, it fully depends on your methodology. Anytime I see bytes being used, it is usually with binary prefixes (multiples of 1024) instead of metric (multiples of 1000), whereas when bits are used it is usually with straight metric system calculations. I see this even when dealing with communications which is weird. In fact even in my advanced networking technologies courses the professors would accept either method as long as you explained how you did it. Even though the binary prefix is MiB or GiB, it is commonly written as MB or GB still, even in textbooks.

So if you have a metric calculation with bits that is 1Gb, that is 186,000,000,000 bits. Whereas when I see it specified in GB it usually applies the prefix meaning 1,073,741,824 bytes. There are tons of calculators out there that do this such as:

http://www.speedguide.net/conversion.php

and if you put in 186000000000 bits, it comes out as 186 Gb and 21.6 GB.
Click to expand...
byteconversion000.png


Google:
https://encrypted.google.com/#sclie....,cf.osb&fp=5ac088929e1bc45b&biw=1280&bih=895

WolframAlpha Computational Search Engine:
http://www.wolframalpha.com/input/?i=186+gigabit+to+gigabyte

Learn to do your math right son.
 
Furthermore, if you want to convert between bit, kilobit, megabit, gigabit, and terabit, you have to multiply or divide by 1024 bytes. ;) 8 bits in 1 byte and 1024 bytes in 1 kilobyte came first. Computers will always work this way because it is what defines their core in how they function; no matter how hard you want to push this ibibibi system, it will never work because that is not how computers work. If it hurts your eyes and brain too much to use base-2, then I've got some bad news for you buddy: you're in the wrong field of work if your career interest is in computers.

Always cross check your math from both sides, and if they don't equal, then somewhere in the math you made a mistake, didn't do something right, or used incorrect logic.

No offense, but this is the way it works with mathematics in reality and with computers.
 
aronesz said:
byteconversion000.png


Google:
https://encrypted.google.com/#sclie....,cf.osb&fp=5ac088929e1bc45b&biw=1280&bih=895

WolframAlpha Computational Search Engine:
http://www.wolframalpha.com/input/?i=186+gigabit+to+gigabyte

Learn to do your math right son.
Click to expand...

umm.. no. The Giga- prefix has two different definitions depending on the context. So a single conversion doesn't give you the right contextual answer. 23.25GB * 2^30 (binary definition of a Gigabyte) = 200Gb (rounded up). 21.65GB * 2^30 = 186Gb (rounded up)
 
And here is a case where the rule of "never argue with a group of newbies" applies. I'm done.
 
aronesz said:
And here is a case where the rule of "never argue with a group of newbies" applies. I'm done.
Click to expand...

You being the newbie, I guess. Bits have been measured with base 10 prefixes since the very beginning of computer science. 1kbit = 1,000 bits, not 1,024. because kilo- mega- giga- etc were defined long before digital computers came around. It wasn't until after bytes started being used (which originally were 6bits, mind you) to measure character capacity that the 1024 multiplier started being used because it was easier to compute and was close enough to 1000 to not make much difference. But if you wanted accuracy, you used bits and the predefined metric system..
 
Use the proper standardized symbols or label context when doing math in relation to computers.

Code: 
186 Gb = 23.25 GB = 21.653 GiB
186 metric (SI) Gb = 23.25 metric (SI) GB = 21.653 binary (ISO-80000) GiB

Note that context changes the meaning of "GB" when standardization is ignored

Code: 
186 Gb (network)  = 23.25 GB (network) = 21.653 GB (disk)
 
sabar said:
Use the proper standardized symbols or label context when doing math in relation to computers.

Code: 
186 Gb = 23.25 GB = 21.653 GiB
186 metric (SI) Gb = 23.25 metric (SI) GB = 21.653 binary (ISO-80000) GiB

Note that context changes the meaning of "GB" when standardization is ignored

Code: 
186 Gb (network)  = 23.25 GB (network) = 21.653 GB (disk)
Click to expand...

Finally, someone gets the whole point of my posts. When you look at transfer speeds at some websites they incorrectly give you speeds determined by the amount of "storage" moved instead of the amount of "bits" transferred. If you want to measure speed based off of storage (which some people argue saying that communication companies are "misleading" them), then you have to do the math I showed. You have to explain how quickly you move 10GB of "storage", which is in GiB and is listed as GB. So that is what the math is showing.

aronesz said:
And here is a case where the rule of "never argue with a group of newbies" applies. I'm done.
Click to expand...

You should have been done before you started incorrectly trying to correct people. And as far as "newbies", you don't even understand the basic principles here and you have only been a member for 60 days, yet you are trying to label others as newbies?
 
On the topic of bits vs. bytes, I wish more manufacturers, authors, etc. would use MiB, GiB, etc. instead of MB, GB, etc. to differentiate power of 2 vs. power of 10, like sabar did.
 
latenite said:
It all doesn't matter because Windows has a problem moving files much faster than 2-3 MB/s on a 1Gbps interface
Click to expand...

um, right....

tested gigabit transfer b/w 2 computers (hooked up to a gigabit Asus RT-N16 tomato)

~115mbyte/s transfer... both are using ramdrives, so no bottleneck there. same speeds when using the SSDs instead...
 
latenite said:
It all doesn't matter because Windows has a problem moving files much faster than 2-3 MB/s on a 1Gbps interface
Click to expand...

LOL

As much as I dislike Windows, I do have to admit you are incorrect on this one, unless you are joking of course. ;)

Regardless of the OS, as said above ^ the maximum transfer rate is about 115MB/s and without overhead is 125MB/s.
 
Red Falcon said:
LOL

As much as I dislike Windows, I do have to admit you are incorrect on this one, unless you are joking of course. ;)

Regardless of the OS, as said above ^ the maximum transfer rate is about 115MB/s and without overhead is 125MB/s.
Click to expand...
No, it'd be 119.21 MB/s, not 125 MB/s. :rolleyes:
 
aronesz said:
No, it'd be 119.21 MB/s, not 125 MB/s. :rolleyes:
Click to expand...

Are you joking or being serious?

1000Mbps / 8 = 125MB/s

It has and always will be this way.

Now fiber is a bit different since there is the 8b/10b conversion.
Is that what you meant?
 
Red Falcon said:
Are you joking or being serious?

1000Mbps / 8 = 125MB/s

It has and always will be this way.

Now fiber is a bit different since there is the 8b/10b conversion.
Is that what you meant?
Click to expand...
So then why isn't 186 Gbps 23.25 GB/s as I have argued?
 
Red Falcon said:
Are you joking or being serious?

1000Mbps / 8 = 125MB/s

It has and always will be this way.

Now fiber is a bit different since there is the 8b/10b conversion.
Is that what you meant?
Click to expand...

Only if you use the standard definition of mega (10^6) rather thean the traditional (but ambiguous) deifintion (2^20). You're both right

aronesz said:
So then why isn't 186 Gbps 23.25 GB/s as I have argued?
Click to expand...

It is if you use the -ibi standard that you detest so much. 23.25GB/s == 21.65GiB/s
 
aronesz said:
So then why isn't 186 Gbps 23.25 GB/s as I have argued?
Click to expand...

If the 186Gbps was using ethernet, then you would be correct.

NoOther is converting ethernet bytes to storage bytes, as shown on that calculator, which also makes sense.

However, he is forgetting about protocol overhead, which is also present in sending data across ethernet or fiber.

So in the end, the actual amount of data sent is 1000Mbps = 125MB/s, though this is not reflected upon the data on the disk due to the protocol overhead.
 
Red Falcon said:
If the 186Gbps was using ethernet, then you would be correct.

NoOther is converting ethernet bytes to storage bytes, as shown on that calculator, which also makes sense.

However, he is forgetting about protocol overhead, which is also present in sending data across ethernet or fiber.

So in the end, the actual amount of data sent is 1000Mbps = 125MB/s, though this is not reflected upon the data on the disk due to the protocol overhead.
Click to expand...

a byte is a byte. 125,000,000 bytes is no different on disk than it was coming through the cable. The difference comes in how you define the prefixes used to represent large numbers of bytes. That's why IEC is pushing (or was) to have a the SI prefixes separated from their binary definitions. Protocol overhead is another matter entirely.
 
ryan_975 said:
a byte is a byte. 125,000,000 bytes is no different on disk than it was coming through the cable. The difference comes in how you define the prefixes used to represent large numbers of bytes. That's why IEC is pushing (or was) to have a the SI prefixes separated from their binary definitions. Protocol overhead is another matter entirely.
Click to expand...

Correct on the byte being a byte part, but you forget about additional packets being added on to each disk-byte of data, thus increasing what is being transferred over ethernet, hence the overhead on ethernet.

1000Mbps = 125MB/s

I can't believe this is even a debate.
It's been this way since ethernet was created.
 
Red Falcon said:
Correct on the byte being a byte part, but you forget about additional packets being added on to each disk-byte of data, thus increasing what is being transferred over ethernet, hence the overhead on ethernet.

1000Mbps = 125MB/s

I can't believe this is even a debate.
It's been this way since ethernet was created.
Click to expand...


As I said before overheard is another matter. 1,000,000,000 bits / 8 bits per byte = 125,000,000 bytes. How much of that data makes it to the disk isn't our concern at this point, 125,000,000 bytes were still transferred.

The debate isn't about how much data was transferred, but what to call the amount that was transferred. Because of the ambiguous use of SI prefixes, some guys (mostly communications folks) will say you transferred 125MB, while others will say you transferred 119.21MB. In both cases they mean that 125,000,000 bytes worth of ethernet overhead and payload came across the cable.
 
ryan_975 said:
As I said before overheard is another matter. 1,000,000,000 bits / 8 bits per byte = 125,000,000 bytes. How much of that data makes it to the disk isn't our concern at this point, 125,000,000 bytes were still transferred.

The debate isn't about how much data was transferred, but what to call the amount that was transferred. Because of the ambiguous use of SI prefixes, some guys (mostly communications folks) will say you transferred 125MB, while others will say you transferred 119.21MB. In both cases they mean that 125,000,000 bytes worth of ethernet overhead and payload came across the cable.
Click to expand...

Ah, that makes sense. Thanks for the clarification.
 
ryan_975 said:
As I said before overheard is another matter. 1,000,000,000 bits / 8 bits per byte = 125,000,000 bytes. How much of that data makes it to the disk isn't our concern at this point, 125,000,000 bytes were still transferred.

The debate isn't about how much data was transferred, but what to call the amount that was transferred. Because of the ambiguous use of SI prefixes, some guys (mostly communications folks) will say you transferred 125MB, while others will say you transferred 119.21MB. In both cases they mean that 125,000,000 bytes worth of ethernet overhead and payload came across the cable.
Click to expand...

And thus this gets back to my original point about the so-called "misleading" information. It is simply a matter of perspective. Some people want to know how fast they can transfer the information (networking), and some people want to know how fast their particular data is moved (storage). And yes in my calculations I didn't fully account for overhead, but different communication systems may have different overhead as well.
 
You must log in or register to reply here.
Back
Top