CNet rides again

[James]

http://reviews.cnet.com/techtrends/0-6014-8-20069949-1.html?tag=st.re.9870989.fs.6014-8-20069949-1

What do you think? Are we obsessed with unnoticeable differences in speed?

 

[NRG = mc²]

I reckon CNet is FOS.

I can certainly tell the difference between a 40Gb D740X and a 80Gb Barracuda IV in quiet mode, the latter being annoyingly slow at times (loading ICQ, virus scan, Acrobat, etc). And they say you might as well get a 5400rpm drive - now if one picks up a Seagate U-series I'm willing to bet my house they'll see a difference compared to just about any 7200 rpm drive (excluding some Seagate offerings).

 

[James]

Heck, CNet could hardly tell the difference between an X15-36LP and a D540X...

 

[James]

Oh, and while it was good to point out the ATA133 interface is to all intents and purposes no faster than ATA100, I think they kind of missed the main point about it, no?

 

[time]

I feel dirty now. James, what possessed you to sully our pristine forum with such drivel?

How about they do the same article on the lack of differences between CPUs? Because Intel is a huge sponsor, that's why. So I guess the fact that you can't discern any difference between a P4 1.6 and a P4 2.53 in desktop applications will remain unreported.

Too bad Seagate and Maxtor didn't spend more on advertising with CNet, eh?

BTW, what's all this "EIDE" crap? Isn't that WD's terminology?

 

[P5-133XL]

Sorry to disagree with the posts here, but I think CNET is correct. For most drives the difference in speed is not noticable except through benchmarking.

With that said, modern High end SCSI is signifigently faster than IDE and Notebook drives are much slower than everything else. Thus there is a difference between the extremes

My general rule of thumb ( for all computer benchmarking) is that it takes a 100% increase in speed to be signifigent productivity improvement. Any difference less than 20% is only detectable by benchmarking and between 20% and 50% improvement is only detectable by the average user. between 50% and 100% is debatable. On this generalized scale most drives are not distinguishable between each other.

Thus I don't recomend upgrading anything unless there is a clear 100% improvement in speed or there is a threshold that needs to be broken like the 70Hz refresh speed of a monitor. It just isn't worth much going from a PIII500 to a PIII 800 or a 15ms Access time to a 12ms Access time. Going from a 5400RPM drive to a 7200RPM drive is noticable and going from a 20ms to a 12ms is noticable. But unless the difference is large, there just isn't enough to justify much price difference

 

[Cliptin]

Sorry to disagree with the posts here, but I think CNET is correct. For most drives the difference in speed is not noticable except through benchmarking.

With that said, modern High end SCSI is signifigently faster than IDE and Notebook drives are much slower than everything else. Thus there is a difference between the extremes

My general rule of thumb ( for all computer benchmarking) is that it takes a 100% increase in speed to be signifigent productivity improvement. Any difference less than 20% is only detectable by benchmarking and between 20% and 50% improvement is only detectable by the average user. between 50% and 100% is debatable. On this generalized scale most drives are not distinguishable between each other.

Thus I don't recomend upgrading anything unless there is a clear 100% improvement in speed or there is a threshold that needs to be broken like the 70Hz refresh speed of a monitor. It just isn't worth much going from a PIII500 to a PIII 800 or a 15ms Access time to a 12ms Access time. Going from a 5400RPM drive to a 7200RPM drive is noticable and going from a 20ms to a 12ms is noticable. But unless the difference is large, there just isn't enough to justify much price difference

Don't forget that upgrading is different from a new purchase. I do agree with you to a certain extent, though.

 

[NRG = mc²]

a 15ms Access time to a 12ms Access time

Isn't this the difference between a good 5400 and a good 7200rpm drive?

Sure, a 5400rpm doesn't neccesarily improve productivity, I'm sure we all could live with 5.4k but its damn annoying, and reduces the percieved snappyness of the system

 

[P5-133XL]

Don't forget that upgrading is different from a new purchase. I do agree with you to a certain extent, though.

Even with a new purchase, how much should you pay to get a device that has an inperceptable speed gain that only shows up in benchmarks? I would say little to none and thus go for the best buy rather than the best performance. There are issues concerning features, service, warrentee, and reliability to be included as factors and those are far more important than performance differences.

Shall I start an arguement? sure - What's the point of overclockability, different ram types, chipsets, even cpu's when the performance differences are less than 20%. Why pay up for top-of-the-line when middle of the road is so fast that one can't tell the difference. As more time goes by, the more I'm inclined to pay attention to price and other factors and less on performance for just about everything associated with computers.

 

[Cliptin]

Don't forget that upgrading is different from a new purchase. I do agree with you to a certain extent, though.

Even with a new purchase, how much should you pay to get a device that has an inperceptable speed gain that only shows up in benchmarks? I would say little to none and thus go for the best buy rather than the best performance. There are issues concerning features, service, warrentee, and reliability to be included as factors and those are far more important than performance differences.

Shall I start an arguement? sure - What's the point of overclockability, different ram types, chipsets, even cpu's when the performance differences are less than 20%. Why pay up for top-of-the-line when middle of the road is so fast that one can't tell the difference. As more time goes by, the more I'm inclined to pay attention to price and other factors and less on performance for just about everything associated with computers.

Don't get me wrong, I agree with you. But as you stated yourself, there is a difference between the X1536 and an OEM 5400RPM drive however slight. It is up to the individual consumer to determine whether or not the different is cost justified. Just like everything else in life. Cars, boats bicycles, appliances....

PS you did notice I'm selling my X15 didn't you? :wink:

 

[Mercutio]

Wow. I'm stunned by the profound ignorance involved in writing that article.
Most especially, I can't believe how dismissive they are of 15krpm SCSI drives, but more than that, the overwhelming message they have seems to be "Buy whatever is cheapest", which I don't think is the right message to present to consumers.

Imagine applying the same logic to cars:
"We're all limited to 55mph anyway, so there's no difference between this Geo Metro and that Porsche 911 except the Porsche costs too much."

 

[James]

Shall I start an arguement? sure - What's the point of overclockability, different ram types, chipsets, even cpu's when the performance differences are less than 20%. Why pay up for top-of-the-line when middle of the road is so fast that one can't tell the difference. As more time goes by, the more I'm inclined to pay attention to price and other factors and less on performance for just about everything associated with computers.

That's a valid point of view nowadays Mark, I don't think you'll get many people here to disagree. But overclocking was useful back when (a) CPUs were much more expensive and (b) a 50% CPU performance increase made a noticeable difference - ie. things were somewhat CPU bound. Now they are for the most part not and I bet you would see little difference (as has been said above) in day to day use of a 1.6GHz and 2.53GHz P4. But there was certainly a world of difference between a 300MHz Celeron A and the same chip at 450MHz.

As a result, I think the biggest determinant of CPU buying today should be price, and indeed I suspect it will remain that way for some time to come. Office applications aren't pushing the performance envelope, games are - and games are increasingly being tuned to run on expensive graphic cards rather than sucking CPU cycles. I wouldn't be surprised if nowadays gamers upgrade graphics cards more often than their CPUs.

 

[Adcadet]

Note what they say about seek time/access time and buffer size. Perhaps they should read up over at SR.


Seek time: This relatively meaningless number describes how long it takes, on average, for a hard drive's read/write heads to find a random track. A better number--which you won't always find listed--is the access time, which takes into account the latency incurred by the head waiting for the platter to spin before a track can be accessed. A drive's access time may help predict performance in server applications, where simultaneous calls from multiple users cause the read/write heads to fly all over the platters. But on the desktop? You'll never spot the difference.

Cache size: Every hard drive has a chunk of memory to hold frequently accessed data--usually 2MB worth. To determine the effects of a larger cache, we couldn't resist testing the new Western Digital Caviar 120GB Special Edition, which comes with an 8MB buffer. The conclusion? You guessed it: no discernible boost over similar 2MB drives in desktop applications. (Fortunately, there's little difference in price, either.)