[thideras]

Originally Posted by Moto7451

It's like asking how many slopes (remember from Algebra, rise over run) does an apple weigh.

I love that example

[gulp35]

You could maybe measure the "Hz" of a HDD by converting the average seek time (or rotational latency) into the # of clock cycles that would pass on your computer (i.e. how many cycles your processor would have to wait if data had to be fetched from the HDD). Though in that case... You would want the smaller number...

[Moto7451]

It really doesn't work like that though. You can't measure the Hz of a hard drive *period.*

What you're talking about is simply latency. Also, the hard drive isn't going to cause everything to come to a stand still, the CPU is still busy doing things like running the OS and software thats already opened. Most programs will load enough of themselves into memory to avoid the CPU stopping for something to be read from the HD.

I'm still sure the reason they tried to equate rotational speed to processor clock speed was to impress on the reader the idea that a faster drive makes for a faster overall system and thats it.

[MadMan007]

I think Moto is getting too caught up in the technicalities of it rather than seeing it as an anology.

[Mr.Guvernment]

Slowest part is the DVD / Cd drive

[thideras]

Originally Posted by Mr.Guvernment

Slowest part is the DVD / Cd drive

Yup, fire up a virtual machine with an ISO as the cd drive. Watch a windows install (moving files, getting drivers) and you will be amazed. It takes about a second or two to load all the drivers (which normally takes minutes on my computer) and copying all the files take a good 15 seconds (again minutes regular).

[Mr.Guvernment]

Will be niec when the media drivers speed up, i am sure HD / Blu-ray may help, but how soon will we see games and OS on Blu-ray disks...

[MadMan007]

Windows 7. More bloat = Blu-Ray! Although I'm sure MS would have preferred HD DVD

[Moto7451]

Originally Posted by MadMan007

I think Moto is getting too caught up in the technicalities of it rather than seeing it as an anology.

lol NO I keep saying that all it is is an analogy .

Originally Posted by Me

If they were trying to equate anything between the two measures, it was just metaphorically, i.e. a 7200RPM drive is like a 2GHz C2D as a 15K RPM drive is to a 4GHz C2D.

....

I'm still sure the reason they tried to equate rotational speed to processor clock speed was to impress on the reader the idea that a faster drive makes for a faster overall system and thats it.

[jason4207]

Rotation can be directly related to up and down motion. The piston head in an engine moves up and down over time, and turns things in a rotational fashion.

*Devil's advocate*

[Moto7451]

Thats only because there's a cam shaft in the engine that converts up down to rotational... hard drives lack this.

Stop trying lol please just everyone just stop trying to figure out what Maximum PC was trying to "measure" in hertz. This thread is a riot lol.

All they wanted to do was make an analogy/comparison/try to give people the idea that faster HD = better.

[jason4207]

Originally Posted by Moto7451

This thread is a riot lol.

My thoughts exactly...I'm just having fun!

*edit...I believe it's the crankshaft, not the camshaft that converts the up-down motion to rotational motion. Someone correct me if I'm wrong. But yes, HDD's do lack crankshafts! LOL!

[JamesXP]

Yes, the cam opens up the valves for the 4 stages of yada.


I don't see HDs exactly being a bottle neck in games etc, they just make load times longer, nothing that slows gameplay down, but in video/sound/etc editing it would make some kind of difference, I do agree such formula is spoon-talk.

[Firestrider]

Well if you look at a computer as a memory hierarchy then yes the hard drive is one of the slowest parts of the system. If you look at it from the top it goes registers (accumulator) -> cache (L1/2/3) -> main memory (RAM) -> flash memory (USB Drive) -> disk storage (Hard Drive) -> tertiary storage - optical disk (DVD-ROM Drive). Every level down there is an increase in latency and a decrease in throughput.

Now I guess you could convert the hard drive into mhz if you consider latency analogous to hertz. Say you have a processor running at 3 Ghz. The registers would have a latency of 1/(3*10^9) seconds or 0.33ns analogous to 3 Ghz. Now say your hard drive has an access time of 5ms; that would be analogous to 15 Mhz.

[thideras]

Originally Posted by Firestrider

Now I guess you could convert the hard drive into mhz if you consider latency analogous to hertz. Say you have a processor running at 3 Ghz. The registers would have a latency of 1/(3*10^9) seconds or 0.33ns analogous to 3 Ghz. Now say your hard drive has an access time of 5ms; that would be analogous to 15 Mhz.

That does not even make sense >.<

All you are proving is that lower latency is better.

Like Moto said, can't compare. Period.

[freakdiablo]

Originally Posted by Firestrider

Well if you look at a computer as a memory hierarchy then yes the hard drive is one of the slowest parts of the system. If you look at it from the top it goes registers (accumulator) -> cache (L1/2/3) -> main memory (RAM) -> flash memory (USB Drive) -> disk storage (Hard Drive) -> tertiary storage - optical disk (DVD-ROM Drive). Every level down there is an increase in latency and a decrease in throughput.

Now I guess you could convert the hard drive into mhz if you consider latency analogous to hertz. Say you have a processor running at 3 Ghz. The registers would have a latency of 1/(3*10^9) seconds or 0.33ns analogous to 3 Ghz. Now say your hard drive has an access time of 5ms; that would be analogous to 15 Mhz.

While I get it, that would only make sense if drive speed was linked to cpu speed.

[Firestrider]

I was just trying to make a correlation and this is probably went through the minds of that PC magazine.

[Moto7451]

I was thinking crankshaft, just used the wrong words. Usually I just call it a cam... cam this cam that. Thankfully I'm not a mechanic .

[UBERCOOL]

Originally Posted by Duner

You guys have to stop thinking pure computer and just general physics.

A hertz is a cylce/second of anything, not just CPU's.

Most North American Household power is supplied as 120V and 60Hz. The AC cycles 60 times/second.

If you take a 7200 rpm drive. That spins 7200 revolutions per minute. That equals 7200 revolutions per 60 seconds or 7200/60 = 120 revolutions per second. In otherwords 120 cycles/second or 120 Hz.

Duner is right. Hz is defined as cycles per second(can be used to measure oscillations but is not limited to that), and in the case of a motor a cycle is considered a rotation. Moto is right in the aspect that you can't really compare Hz from different things because while it is the same unit, they are different measurements. However, you can easily convert from RPMs to Hz using Duner's method.

[bLack0ut]

Technically, if you plot the x/y graph of a rotation over the independent variable time, there is a clear sinusoidal wave (and its resulting frequency). So you just have to isolate the x or the y function and relate it to time.

This could be relevant in the sense that data is usually stored at a specific point, and the hard drive will have to complete a (partial) revolution in order to reach the next point.

Disclaimer: I know next-to-nothing about how hard drives work :P