hard drive...

[AND1]

Hi I already have a 120 GB hard drive and want to add another one. I dont really know much about them but i was wondering if any of you know which is the best. I will use it for ripping my dvd movies onto. I'm thinking about getting a maxtor? Have any ideas?

[mantisgeek]

Two that springs to mind are the Western Digital Caviar WB1200JB (or the WB1200BB) or the IBM Deskstar 120GXP. For the WD Caviar, the difference between the JB and the BB is the size of the buffer; BB having 2MB and JB having 8MB buffers. I have the WB1200BB and am very happy with it. The IBM is slightly less expensive than the WD but a couple of my friends have had problems with IBM drives after a couple of years and based on this (albeit limited) experience, I recommend you go for the WD instead.

[thehutt]

I would aviod the Maxtor....I have had problems with them and I have heard bad things from others as well.

I would go with Western Digital....and the faster 7200 rpm as well.

[HillJack]

Normally, I would say IBM, but steer clear of the 120GXP - they are notorious for short life spans. Their 60 and 80giggers are nice though (I own 3 of them).

I agree with the others though, WD is an excellent brand. Maxtor, well, they're just as reliable as Memorex CD-R's

[fakesky]

No problems with my (2) ibm 10GXP

[Nelson37]

I have installed literally HUNDREDS of Hard drives. ALL major manufacturers have had some bad models, MOST are fine. WD are probably my favorite, although their six or eight gig model had a problem with the heads falling off. Maxtor has purchased Quantum, Quantum drives seem to have higher than normal failure rates. Seagate has had several bad models, we had six of eight high-end SCSI drives fail within 30 days. IBM deskstar have had some high failure rates.

ALL major brands carry 3-year warranty now. NEVER buy a brand-new model, wait till it has been in production a year or so.

Remember this is a numbers game. If ONE person has either success or failure with ONE drive from ONE manufacturer, this means nothing. Also, many systems integrators purchase refurbished drives, which are more likely to fail - so that make and model of drive has failed TWICE.

Western Digital and Maxtor support their products with full replacement warranty, w/Maxtor you must specify NO QUANTUM. 7200 RPM or faster, UDMA 66 or better, make sure you get the good cables for UDMA.

[Mirror_Image]

Nelson37

You are absolutely correct in hinting at bad statistical sampling and analysis being the cause of people drawing inaccurate conclusions.

(I wrote this some time ago):

---------------------------------------------
We consider 3 fictional widget-manufacturing companies, and these 3 make all the widgets on the market.

The ABC company makes 75% of all widgets sold, the QRS company makes 24% of the widgets sold, and the XYZ company makes 1% of the widgets sold. In other words, for every 100 widgets that are sold to the public, 75 are ABC, 24 are QRS, and 1 is XYZ.

ABC widgets are very well-known, QRS widgets are fairly well known, and not many people have heard of XYZ widgets. because they take advantage of the phenomenon sometimes known as "ecomomy of scale", ABC costs are lower, and their widgets are the cheapest at $100; for the same reasons, QRS widgets cost a bit more at $110, while XYZ widgets cost a pricey $150.

On a widget-enthusiast internet message board, we start a thread titled "If you have had trouble with any widgets, post a reply and tell us about it." It would be reasonable to assume that any large sample of the board's users will roughly mirror the 75/24/1 breakdown of the 3 manufacturers' sales.

100 people who have had problems with their widgets respond to our thread. They posts are 50%/40%/10% ABC/QRS/XYZ. If we didn't know the sales figures and the market share of the 3 companies, we might wrongly deduce that XYZ was much more reliable than QRS, and QRS was slighly more reliable than ABC. We might look at their pricing 100/110/150, and also wrongly deduce that "if you want a good widget, you have to pay more".

If we look closer and back out (admittedly unreliable, but we'll do it anyway) defect rates for each manufacturer, we see that XYZ has the highest defect rate (as well as being the most expensive), with QRS's defect rate being much lower, with ABC's widgets being the most reliable. Quite a shock, because we thought XYZ was better, and thus worth the cash - and we were wrong.

In any case, if 1,000,000 widgets are sold each year, a flawed study of 100 users is almost laughable. Substitute the peripheral of your choice for "widget".
---------------------------------------

[Nelson37]

Another factor to consider is that some manufacturers sell more drives for user-installation, one would expect a higher failure rate for these rather than professionally-installed drives.

These same analysis techniques extend to motherboards and capture cards (in fact to damn near everything).

To anyone seeking to analyse failure rates, you need LARGE NUMBERS of reports to draw any conclusions, and also some market share info as stated above.

Mirror_Image - excellent analogy, it just annoys me that so many people don't THINK. I constantly have people ask me, after I have told them to save their file, "what should I call it?" My stock answer is "Well, you could call it BOB", and then, to their puzzled look, I elaborate "How the hell should I know what YOU should call YOUR file?"