Tech Things You Dont Know But Are Too Afraid To Ask.

[Godlygamer23]

Quote from Wikipedia article on Chemical compounds: "Chemical compounds can be molecular compounds held together by covalent bonds, salts held together by ionic bonds, intermetallic compounds held together by metallic bonds, or complexes held together by coordinate covalent bonds."

 

http://en.wikipedia.org/wiki/Chemical_compound

[Vitalius]

@Godlygamer23

@Vitalius

 

http://en.wikipedia.org/wiki/Molecule

 

 Complete with 6 sources.

Hmm, cool.

What do you call a group of atoms connected by metallic bonds then?

[Vitalius]

Quote from Wikipedia article on Chemical compounds: "Chemical compounds can be molecular compounds held together by covalent bonds, salts held together by ionic bonds, intermetallic compounds held together by metallic bonds, or complexes held together by coordinate covalent bonds."

 

http://en.wikipedia.org/wiki/Chemical_compound

You do realize that doesn't cover silicon right? 

"intermetallic" As in, using more than 1 type of metal atom. Silicon just uses silicon. That's the idea anyway. Not sure if it's true.

[Godlygamer23]

You do realize that doesn't cover silicon right? 

"intermetallic" As in, using more than 1 type of metal atom. Silicon just uses silicon. That's the idea anyway. Not sure if it's true.

Ghost said silicone. Not silicon.

[Godlygamer23]

Besides, silicone is defined as (Wikipedia) "inert, synthetic compounds with a variety of forms and uses."

https://en.wikipedia.org/wiki/Silicone

[Vitalius]

Ghost said silicone. Not silicon.

Sorry, mixed up words that look the same, mostly.

Not seeing a difference in that though. Still purely silicone.

[Carter]

Hmm, cool.

What do you call a group of atoms connected by metallic bonds then?

A metal?

(I'm not being sarcastic btw)

[Ghost]

Well I'm not going to bother going into it but I was partially wrong. Silicon is a semi-metal which has "different" properties so I have no idea about its bonding. All I know is that it is a giant covalent structure but I am still uncertain of it being called a molecule due to its size and that it has "different" bonding. If someone who has actually been taught about this could speak up I would be interested.

 

__________________________________

 

That was a typo on my part with the silicone. Silicon is used in electronics.

[Godlygamer23]

Well I'm not going to bother going into it but I was partially wrong. Silicon is a semi-metal which has "different" properties so I have no idea about its bonding. All I know is that it is a giant covalent structure but I am still uncertain of it being called a molecule due to its size and that it has "different" bonding. If someone who has actually been taught about this could speak up I would be interested.

 

__________________________________

 

That was a typo on my part with the silicone. Silicon is used in electronics.

So you're referring to silicon then?

[Vitalius]

A metal?

(I'm not being sarcastic btw)

I feel like that's just.... too easy. My experience with Chemistry (classes) tells me it's more complicated than that. You are likely right though.

Anyway,

 

I asked this in a different way earlier, but hyper threading is basically taking your CPU cores and using them for one thread when the thread it's working on is on hold as it fetches the data from the RAM because the first time it did it, it had an error (or something like that). That's at least how I understand how hyper threading works, and I could be wrong, but assuming I'm right, what kind of effect does higher speed RAM have on hyper threading since the RAM can react faster?

... I probably butchered that HT explanation or am wrong. 

[Glenwing]

Yes.  Whenever the cores are idle (due to anything, waiting on RAM, cache miss, etc.) it uses the idle cycles to do something else.

[Godlygamer23]

Hyper threading. May not be 100% correct.

[Ghost]

So you're referring to silicon then?

Well yes. That was the point of this discussion. The original argument was that the CPU traces couldn't be less than 1 atom thick. My argument is that they could be.

 

Anyway back to silicon it IS a molecule. Source: http://inventors.about.com/library/inventors/blsolar4.htm

 

But then at the end this: "This regular, fixed arrangement of silicon atoms is known as the "crystal lattice."" So now I am really confused.

[Glenwing]

... But what makes RAMDisks awesome is the bandwidth. The 1-3 GB/s transfer speeds. I'm pretty sure 2133 memory gets the same bandwidth as 1333 memory, it just responds (CL) and acts (the actual speed) faster? I could be wrong, but I don't think I am.

 

I can understand them not being 1 atom wide as they are very likely made up of molecules, but I believe they can be 1 molecule wide as long as they are in a contained environment.

 

If they were a molecule wide they would just be a molecule.  A conventional transistor is not a molecule, but a construction consisting of two or more gates, a source, and a drain, unless you are going to talk about using electrons in atoms as bits (quantum computing).

[Vitalius]

If they were a molecule wide they would just be a molecule.  A conventional transistor is not a molecule, but a construction consisting of two or more gates, a source, and a drain, unless you are going to talk about using electrons in atoms as bits (quantum computing).

Well, no. Didn't know that. Was imagining a transistor as the pins on a CPU. I fail.

And, 

Yes.  Whenever the cores are idle (due to anything, waiting on RAM, cache miss, etc.) it uses the idle cycles to do something else.

So it's not just specifically RAM.

But, how would faster RAM (such as 1333 VS 2333) affect that? Or have you ever heard of anyone testing it?

[Glenwing]

Well, no. Didn't know that. Was imagining a transistor as the pins on a CPU. I fail.

And, 

So it's not just specifically RAM.

But, how would faster RAM (such as 1333 VS 2333) affect that? Or have you ever heard of anyone testing it?

 

Haven't heard of tests but I imagine it wouldn't affect much given that the latency doesn't really change (~10ns).  Just the bandwidth.

[Vitalius]

Haven't heard of tests but I imagine it wouldn't affect much given that the latency doesn't really change (~10ns).  Just the bandwidth.

Wait. I thought the bandwidth didn't change from RAM speeds. Just the latency/reaction time. 

... I don't understand RAM speeds anymore (or didn't to begin with). What is the actual difference between 1333 and 2133 RAM then? What do the numbers represent in real world terms (not just "Megahertz")?

[Godlygamer23]

Clock speeds mean nothing today. It goes for both RAM and CPUs.

[Vitalius]

Clock speeds mean nothing today. It goes for both RAM and CPUs.

Except when you are using an APU or video editing. Otherwise you are right.

[Glenwing]

Wait. I thought the bandwidth didn't change from RAM speeds. Just the latency/reaction time. 

... I don't understand RAM speeds anymore (or didn't to begin with). What is the actual difference between 1333 and 2133 RAM then? What do the numbers represent in real world terms (not just "Megahertz")?

 

The numbers represent how many cycles the RAM performs each second.  1333, 1600, etc. are "effective MegaHertz" as DDR-DDR3 performs multiple calculations per cycle.

 

CAS Latency is the time it takes for the RAM to get back to the CPU after the CPU makes a request.  It's measured in cycles (CL9 = it takes 9 cycles).  In reality the latency doesn't really change much; it takes the same amount of time within a small boundary, but of course since it's measured in cycles, as the bandwidth increases, the number of cycles the RAM performs in that amount of time increases, so the latency "increases".

[Vitalius]

The numbers represent how many cycles the RAM performs each second.  1333, 1600, etc. are "effective MegaHertz" as DDR-DDR3 performs multiple calculations per cycle.

 

CAS Latency is the time it takes for the RAM to get back to the CPU after the CPU makes a request.  It's measured in cycles (CL9 = it takes 9 cycles).  In reality the latency doesn't really change much; it takes the same amount of time within a small boundary, but of course since it's measured in cycles, as the bandwidth increases, the number of cycles the RAM performs in that amount of time increases, so the latency "increases".

Cool, thanks.

What is a cycle defined as? The time it takes the CPU to do 1 operation?

[FireFox]

Cool, thanks.

What is a cycle defined as? The time it takes the CPU to do 1 operation?

 

I thought a cycle was from 0 to 1 or 1 to 0 or something

[Godlygamer23]

The definition of cycle "the smallest interval of time required to complete an operation in a computer." (Dictionary.com)

[Glenwing]

Cool, thanks.

What is a cycle defined as? The time it takes the CPU to do 1 operation?

 

Difficult to explain specifically.  Cycle is just a repeated action.  A monitor refresh, RAM operation, CPU operation... depends on which component you are talking about.

[FireFox]

Difficult to explain specifically.  Cycle is just a repeated action.  A monitor refresh, RAM operation, CPU operation... depends on which component you are talking about.

 

Let's say CPU to make it a little simpiler