Every time I've asked about the performance difference between a 7200 RPM drive and a 5400 RPM drive, no one thinks it would make a big difference.

 

Lately the 7200's seem to be substantially more expensive and I came across a deal on WD60EFZX drives (5640) over the holidays ($104 ea) so I picked two up.  

 

After running 2 pre-clear cycles I noticed on drive took 18 minutes longer than the other in each cycle.  I bench marked the drives and they are notably slower than any of the 7200 RPM drives in my array. 

 

I'll be arguing with myself for a while on whether the savings was worth the speed (~ $60 per drive).  It really only matters during a parity check.  This will actually slow my array down because the speed will be more limited once the array passes the 3TB mark and the slower drives drop off.

Edited December 24, 20214 yr by TODDLT

Depends on you focus on performance or cost. For me, cost most important especially for high capacity disk, and slow RPM disk also a bit easy for cooling.

8 hours ago, TODDLT said:

Every time I've asked about the performance difference between a 7200 RPM drive and a 5400 RPM drive, no one thinks it would make a big difference.

It depends on the platter density.

 

There are some 7200 RPM drives that have less throughput than some 5400 RPM drives.

 

IIRC, what was said was that the RPM is only one factor, you can't determine performance solely by spin speed.

 

If you have 2 identically sized drive models with the same number of platters and heads, then yes, RPM is the determining factor. If you have a 4TB drive with 2 platters at 5400RPM vs a 4TB 5 platter at 7200RPM, the 5400 will have a better throughput number and much better power and heat dissipation.

 

Since drive manufactures generally don't publish platter and head counts, you must look at real benchmarks to determine performance. You can't just look at RPM.

I always buy 7200 speed drives. They are most likely to be faster than the 5400 drives, and most of the time have more cache which does make a difference. I love that speedtest docker myself and I see my 3 5400 drives just need to go! Wish I had the cash to replace them all then the curve on my speedtest report would look some much better.