who is your audience for this project? I think you could easily make this document work well for people with low-mid level hardware experience. Also, is this intended for people who are interested in buying a new drive how is it going to influence their purchasing decisions?
However, as it stands, you don't seem to be explaining some key things that should be stated for this audience.... Currently, I think you need atleast a good understanding of HDD's to get some of the finer points of your test.
(1) I would explain why you're not listing capacity in the beginning. Explain why the capacity of the individual platters is more important and why it has more of an effect on the performance of a drive...
(2) You might also want to explain the technical things that affect the HDD speed alittle more... you say that "Areal (platter) density is good because we can get more data on and off of the drive faster."
but you don't say what areal densitiy is.... and you don't say that you want more density vs less. and that areal densitiy is related to platter size.
you also say "Larger buffers are good because as with cache on a processor, having the data in a buffer (or cache) is much faster than actually have to seek and read it. "
this is pretty good... but you might want to say that larger buffers are better because they (1) hold recently used data so that if it needs to be retrieved again it can be done so without any mechanical slowdown and (2) provides more oppertunity for read-ahead caching which should also help reduce mechanical slowdowns...
The difference is that in one you're simply saying that a large cache is faster, in the other you're explaining why it is faster and describing it's function.
You don't seem to mention the fact that higher rotational speed leads to faster access times (the only increase in IDE access times we've had in about 5 years was the jump to 7200RPM)
In the end, you add a AMD AthlonXP 1800+ to the mix... You don't seem to state why, you seem to say that you did these tests to see by "how much" the 8MB cahce improves speed... but you dont state this in your conclusion... such as 'The WD's 8mb cache seems to give the same improvement as upgrading to a 300mHz faster processor in business situations.'
stating that "more impact on performance than the jump from 20GB to 40GB platters" is good, but not sufficient for me.
However, I gather that this data may be more useful if you used it to compare two different situations.... someone is buying a PC.. they can choose between the older HDD and a faster processor, or a slower processor and a fast disk.... which to get? well, now they can have you tell em.
on a last note... I don't particualrly care for the graphs.... I would rather eliminate orange because on some monitors the orange square and red square are hard to tell apart on some graphs... Although... I am all in favor of going to a more "prety" graph if you can make em.
well, the design of the site looks good, the data looks good, page loads fast... you've obviously spent some time... I'd like to see your test methods, but they look like they are coming after the experiment... I hope you were consistent.... It also seems fairly well written.. the only thing that stands out to me is the "what we have also seen" repetition near the end.
Overall, I like it, and see that people are probably going to use this or atleast be interested in it.
I'll post again later - had to go shopping with my wife and bub
