Linus, I'm a high school teacher teaching an A+ Certification class, and your videos have been instrumental in teaching my students! Thank you so so much for them!!
I had a class in high school that taught building computers and electronics, but only recently realized how valuable an A+ Certification class could have been. I hope your still teaching A+ because it should be a standard class in all high schools.
Here's your calculation for internet speed, if the ISP says for example, you have a package which has 25MB/s you just multiply the number with 1024, so in this case 25 x 1024 = 25.600, then just divide that number with 8. The answer should be your top speed in this case 25600 / 8 = 3.200, so it would be about 3.2MB/s 'real' download speed. In many cases the calculation was really accurate.
No, not with 1024, you need to multiply it with 1000. A big confusion because many OSs like Windows can't (they can, but they don't do it) seperate those units. 1 Byte 1 Kilobyte = 1000 Bytes 1 Kibibyte = 1024 Bytes 1 Megabyte = 1000 Kilobytes = 1000000 Bytes 1 Mebibyte = 1024 Kilobytes = 1048576 Bytes and so on. E. g. Linux does that. :)
ManicRobot, depends on the standard you use, IEC or JEBEC. As far as i am aware almost everyone use JEBEC whether they know it or not and the industry uses IEC when they want to screw over the customer with that increasingly bigger getting 1000/1024 difference when calculating speeds or storage.
The JEDEC misuse of SI prefixes should never have been sanctioned. SI prefixes long predate the computing industry and their core purpose is for universal standardisation. On that basis, their defilement by JEDEC is a travesty. The only correct choice for prefixes which represent powers of 2 is the IEC set, which is completely unambiguous and perfectly complements the SI prefixes by being a similar universal standard (JEDEC's transgressions notwithstanding).
Actually the amount of bits needed to store a single character depends on the encoding of the text. ASCII was 7-bit long in its early times, BCD was 6. UTF-16 uses 16 bits, UTF-32... well that's obvious. Also, UTF-8 and UTF-16 are of variable width, but that's a longer story.
In my country they're actually giving me faster internet than the contract I signed, (I'm paying 50 Mbps and I'm downloading at 120 Mbps (98 Mbps since my Network card is onboard)) I'm paying what I'm supposed to pay...
Computers have used 1024 bytes = 1 kilobyte ever since they were invented, because computers count in binary (powers of 2), not decimal (powers of 10). It is only a few recent versions of Linux and Mac OS that have decided to change it to 1000 bytes = 1 kilobyte.
It is physically impossible to manufacture RAM in multiples of 1000 bytes. It has to be made in binary multiples. So therefore it makes perfect sense for the prefixes to be binary as well. Otherwise the Commodore 64 would have to be called the "Commodore 65.536"!
There are 8 bits to a Byte. He said to work out the speed of something (usually given in Megabits/second) in MegaBytes/second, you divide the Megabits speed by 10 to get an approximate MegaBytes speed. His example was 100Mb(its)/s would be roughly 10MB(ytes)/s, when it's closer to 12MB/s. You would only really need to use this to guesstimate how long a file (given in Bytes) would take to transfer at a known speed (given in bits/second). Hope that helped.
UNIX and its variants (AIX, Solaris, etc.) use 1024 bytes = 1 kilobyte too, not just Windows. It is only recent "politically correct" versions of Linux and Mac OS that have changed to decimal. So the vast majority of operating systems in use in the world use binary.
Why doesn't tpg just say 3mbs instead of 24 m bits wtf cares but they say that as misleading advertising because 3 sounds like shit which it is so they say a large number of small scale amounts to sound better. I cant watch a 2k 1440 p without interruptions and buffering. I can only watch 1080p without problems. Ergo its still shit speed in Australia.
The term "byte" itself is inherently binary -- one byte being a group of 8 binary digits (bits). But the hard drive manufacturers ignored that and started using decimal counting of bytes in order to make their drives seem bigger than they really are -- so a "1 (decimal) terabyte" drive is actually only 930 (binary) gigabytes. Seagate uses the terms "binary gigabytes" and "binary terabytes" on their web site FAQ, but still advertises their drives in decimal capacities.
It depends on what you are using your sound for. If it's just for music, it may sound just fine. Once you start gaming or music recording in pro tools, you need a sound card that has good quality because you will be able to pick up sound you either didnt notice before, or in the case of music editing, pick out bad sounding notes to edit out later.
btw the 8086 is the 6 bit proc that all "ibm compatible" procs are based on, thus "x86", and it is a 6bit proc. they came out with the 8088 the next year (it had a math co-processor) that was 8 bit.
1gigabit in an exact conversion is equal to 125MBytes, if you have data in bits just divide it by 8 and you will get the EXACT conversion to megabytes. the reason people use dividing by 10 is because its easy to divide by 10 and its pretty close, enough for an aproximation :)
That's why you would divide by 10 instead of 8, as it's an easier number to divide by. It's not accurate, and is only a really rough estimate but it at least gets you roughly in the right ballpark area. Single digit division isn't too difficult, whereas long division (with double or more digits for both sides of the division) can be a little taxing for some. I think he only gave it as a quick and somewhat accurate technique to work out how long a file would take, but most things tell you anyway.
Hard drives use 512 byte sectors, which is half of 1024 bytes. That's why Windows uses 1 K = 1024 bytes, because that way everything counts up evenly (2 sectors = 1 K) and there is no rounding error.
Depends on how your OS handles inodes and if it uses MBs or MiBs (sometimes it doesn't differentiate the two, and does not state which one it actually uses) to tell filesize
I got a 100Mbps Plan by my Internet Provider and he delivers it.... How is that possible you ask, Linus? Well, he is just providing everyone with 10% more than the contract says... 50 get 55, 100 get 110 and 150 will get 165 Mb/s :) Like your Vids very much.... please keep on :)
Split the binary number into nibbles (add more 0s to the left side if it's not a multiple of four), convert each nibble into denary. Each denary number represents a value, if the nibble is 0-9 then it's 0-9, however if it's 10-15 each number is represented by a letter e.g. A is 10, B is 11 etc. For example 10110111 is represented by B7. Hexadecimal is used to represent colours or IPv6 now we are running out of IP addresses.
People have actually sued hard drive manufacturers for using 1k = 1000 because it misleads them into thinking the drive is bigger than it actually is. And using 1k = 1000 doesn't even make sense, because hard drives are manufactured using 512 byte sectors, and 512 is half of 1024. So with 1k = 1024 everything counts up nice and evenly, while if you use k = 1000 you always end up with rounding errors.
The prefixes themselves are not metric. They're just based on Greek words, and are used in different contexts -- for example in electronics, such as kilovolts or megaohms. And the USA doesn't use the metric system in daily life, and even Canada is only partially metric, so that really has nothing to do with it anyway.
Giga, Mega, Kilo etc. are standard metric prefixes with an unambiguous meaning: 1000 of the previous order of magnitude. They are used for plenty of other things besides data storage. 1 Gigawatt = 1000 Megawatts, and 1GHz = 1000MHz. The binary orders of magnitude have their own names. 1GiB = 1024MiB.
Hard drives are advertised with SI units, while Windows gives it to you with binary prefixes. As an example 1 GigaByte is 1,000,000,000 or 10^9 bytes while 1 GibiByte is 1,073,741,824, or 2^20 bytes.
Thanks Linus, thanks for making this video so I can tell people to watch it on repeat instead of trying to explain this simple concept over and over again and try to explain to me that I'm wrong to think that a Gigabit connection can't transfer 1000MegaBytes per second and is faster than SSDs but rather it can only transfer 128MegaBytes per second...
Everyone should know this . Some people with 10Mb/s Internet speeds go to Twitch and watch Livestreams on 3.5MB/sec and complain that the stream lags on their 10Mb/s internet a 3500KB/sec livestream needs atleast a 28Mb/sec internet to watch it without lag , and prefer a faster one cose its hard to hit your max even with a 30Mb/sec
No. Kilo, Mega, Giga are used for plenty of things besides data storage, each prefix mean 1000 of the previous order of magnitude. Even within other areas of computers, 1GHz = 1000MHz. Orders of 1024 have their own names (KiB, MiB, GiB). However Windows has not adopted them yet.
Thanks Linus! I was going to do it xD my list of topics I want to cover has been slowly dwindling since Techquickie started lol. Not that I've actually done anything at all. Need Summer to get here... lol
Metric prefixes (kilo, mega, giga) mean 1000 of the previous order of magnitude, as they do with everything else. The orders of 1024 have their own name, and Windows is the only OS that hasn't fixed their calculation yet in light of the creation of a dedicated set of prefixes for them (the standard is from 1999)
But even Seagate does not use your made-up words "gibibytes" or "tebibytes". On their web site they refer to it as "binary gigabytes" and "binary terabytes". Look up "Why does my hard drive report less capacity than indicated on the drive's label?" in Seagate's Knowledge Base.
Agreed. I suspect because the average joe-blow won't know the difference between bits and bytes, and for download speeds the bigger the better. You're not going to survive if you advertise 20MB/s compared to your competitors advertising 160Mb/s
You have it backwards. In the field of computers, the prefix kilo is commonly accepted to mean 1024. This is well proven fact and any attempt to claim otherwise is false revisionist history. That's why I say they should have invented new words for decimal prefixes (such as kidebytes = decimal kilobytes, medebytes = decial megabytes, etc.) instead of trying to make up new words for the binary prefixes that have been used and accepted ever since computers were first invented.
Actually, the "byte" originally meant the bandwidth of the CPU. If this rule had been applied properly, the max RAM of the 16-bit PC would have been 320kB, however 640kB sounds better, so now the CPU bandwidth is called a "word". You may like to touch on HDD manufacturers who inflate their numbers by eliminating the computing convention of 1024 and advertising their capacities based on 1000. That is one that really screws with the heads of consumers.
he may have the bB nomenclature wrong, but he has a very valid point, where 1000 gigs is considered a Terabyte and 1000 bytes can be considered a K...Damn marketing versus physical reality of hard drives.
Both measurements are correct. The difference is that Windows measures the binary amount of bytes. The other systems measure the decimal amount (grammer might not be correct).
Star Wars: Legacy of the Force series. It's read by Marc Thompson, who does all the voices of all the characters really well, including Han, Leia, Luke, Lando, and others. A very good listen I would recommend to just about anyone, but especially a SW fan. The first one is called Betrayal, by Aaron Allston. They are abridged, and are about Luke and his struggle to figure out who or what is threatening the New Republic after seeing a vision of a dark hooded figure. Anyway, amazingness ensues. :)
How about a video about the difference between the cheaper and more expensive types of video capture equipment? As someone who is just starting to record games, I haven't acquired the experience yet to know if I'm doing my hobby any justice by spending hundreds more on the equipment versus going with something under a hundred bucks. TV and/or PC please. I know it's kind of a tall order for a quickie, but thanks anyway. :)
Yeah (sort of, 1000 kilobytes = 1 megabyte and 1000 megabytes = 1 gigabyte). Not easy to convert, I usually just break it down to the byte level. 1 Kilobyte is 1000^1 bytes, and 1 Gibibyte is 1024^3 bytes. Divide a Gibibyte by a Kilobyte to find out.
Kibibyte / mebibyte / etc. are just made-up Wikipedia words. Nobody actually uses them in real life. People in the computer business are aware that Windows reports RAM and drive space in binary mega/gigabytes (orders of 1024), while the drive manufacturers use decimal giga/terabytes (orders of 1000) to make their drives sound bigger than they actually are. Once you learn that, the context makes it clear whether giga = 1024 or giga = 1000. No need to invent new made-up words to describe it.
I thought the audible thing was just fine, it wasn't intrusive, could be skipped and you made the whole thing interactive by asking us to recommend a book. I think that's awesome. It's a service I use anyway, so I suppose I'm biased but still, making advertising worth listening to has to be something.
He was talking about trying to guess internet speed from a plan. Yes, the correct speed is 12.5 (Dividing by 8) but you will almost never get the claimed internet speed from your ISP. So he says to divide it by 10 to get a VERY approximate estimation.
Tecnically, it's 3.3582324981689453125 MB, a little under 3.5MB, but I'm sure you already know that. lol. Formatting a HD takes out about 9.0329489360253016153971354166670% So if you had a 1TB hard drive, you wouldn't get 1TB after it's formated, it would be something like 931.502602895101 GB
Actually, hard drive vendors nowadays almost always use powers of 10 for measuring storage space (you've probably seen the small-print "1 GB = 1 billion bytes") while Windows uses powers of 2. So 1 TB on a hard drive = 1 trillion bytes. One trillion bytes in Windows equals 931.323 GB. You can verify this by typing "1 TB to GiB" into Google or WolframAlpha or DuckDuckGo or even Bing; "GiB" is the abbreviation for gibibytes, which is just a fancy term for gigabytes in powers of 2.
Kilo means 10^3, Mega means 10^6. These are standardized metric prefixes. We have other terms for the binary orders of magnitude, just that the general public has no knowledge of them...
single rail = normal PSU's multi rail = could be explained as having 2 or more PSU's inside one box. Single rail is easier to set up and more reliable Multi rail gives a better "Price to Watt" value but you have to make sure each rail is connected to an even amount of power draws or else the psu will be inefficient and can even get damaged.
I have a 12 Mb/s connection and almost always get 1,5 MB/s, so I guess it's the real thing they advertise (or maybe they give me little more than 12, so I normally get 1,5 MB/s, because sometime I get even more)
The "world at large" is irrelevant! That's like saying the English language must be changed because it doesn't follow the grammar and spelling of the other languages of the "world at large". *Computers* are what's relevant here, and computers have always used 1 K = 1024 bytes ever since they were invented. It is only very recently that some people have decided to try to change that.
He adjusted it to ten because of network protocols, errors, and inefficiencies in the lines that can potentially cause slow downs. Plus it makes the math so much easier
You've got it backwards. 1 Kilobyte = 1000 Bytes is the correct one. These are standardized metric prefixes, their meaning is unambiguous: 1000 of the previous order of magnitude (1 GigaWatt = 1000 MegaWatts, 1GHz = 1000MHz). The orders of 1024 have their own names. You'll notice the "missing HDD capacity" thing only happens in Windows. OS X and Ubuntu and every other OS does it right (either by changing the displayed units or changing their formula for calculation).
do a video on (might be longer) how to shop for compatible parts. to do this use a screen recorder and go through what it would look like to build a pc from scratch by ordering that parts. this will teach ppl about value, compatibility and overall knowhow!
it's technology that emits more cores than physical cpu cores to the system. if you have 6 core cpu with hyperthreading system detects it as a 8 core cpu... hope that helped.
I am just finishing up a great audiobook on Audible, Hitman, The untold story of Johnny Martorano. Its about Whitey Bulger's enforcer. It's great because the narrator has a great Boston accent. Check it out for sure, It's eye opening. Book is written by Howie Carr
When someone says kB they might mean 1000 Bytes or 1024 Bytes. The technically correct way is to use kB for 1000 bytes & KB (or KiB) for 1024 bytes (but no one ever does it). 1B = 8b (1 Byte = 8 bits) 1kB = 1000 Bytes 1KB = 1024 Bytes = 1KiB The latest convention is to use "kibi, mebi, gibi, etc., to unambiguously denote powers of 1024." en.wikipedia.org/wiki/Kilobyte
I don't know if you're even remotely still interested in a good 'just-for-fun' audible book but I would suggest; The Last Hunter: by Jeremy Robinson. It is a 5 part series that my wife and I have listened to at least a dozen times in less than a year.
Don't you know how file cluster sizes work? Even if you save a 1 byte file, Windows has to save at least one whole cluster to disk, so it will take up 512 bytes of disk space on a FAT32 drive, or 4096 bytes on an NTFS drive. And Windows math is NOT different -- Windows math follows the STANDARD by which ALL operating systems used up until recently, including Mac OS, Linux, UNIX, AIX, etc. It is only within the last few years that Mac OS and Linux changed their math, while the rest have not!
bulldozer is very similar, somewhat better attempt. In a hyper threaded cpu Intel adds some additional processing resources to handle 2 threads per core (nearly 2 full pipelines) to increase performance with a negligible increase in resources. Bulldozed had 2 complete cores, but shared Cache (memory) between them. In this way, a 4 core bulldozer would best a hyper threaded dual core from Intel, but lose to a non hyper threaded Intel quad core (all circumstances presume all cores loaded).
Well not I can stop humiliating myself by complaining to my internet provider. How the fuck was I competent enough to build a PC if I can't even grasp this.
Personally, I think they should do away with the "bits" for internet speeds (as well as SATA speeds!) but the marketing people like the larger numbers. Besides, file sizes are mostly in "Bytes"... so it makes more sense to use MB/s for transfer speeds throughout. To have it in Mb/s (Megabits/sec) is almost as far as using "miles for distance and kilometers/hr for speed".
A "degree" isn't a unit. A degree means a step. "Degrees" is just a shorthand, and yes by region it can vary, but you can always specify, as people do all the time when speaking in an official capacity, they say degrees Fahrenheit or degrees Celcius, not simply degrees. No one thinks of "degrees" as a standalone unit, just like people saying "10 Kilos" to refer to either kilograms or kilometers. But "Kilo" itself isn't an actual unit with 2 definitions. Just as "degree" isn't. Just shorthand.
You obviously don't understand what I was saying. He is saying divide by 10 because that's what you can EXPECT, to get. Obviously its actually divide by 8 but that would be the best speed at which you could possible get where actually you would most likely get a slower speed, so for ease instead of divide by 8 divide by 10. Make sure you understand the comment before trying to correct it.
