Angrymarshmello
Lifer
Before those of you say I'm completely off base, a 5400 rpm drive is faster in some ways than a 7200 rpm drive. I just don't know exactly in what ways and would like to know exactly in what ways it is. Thanks!
-
We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.
In what instances are 5400 rpm drives faster than 7200?
- Thread starter Angrymarshmello
- Start date
Hulk
Diamond Member
I'll just answer briefly since this can become involved.
One way is if you are comparing maximum transfer rates and the areal density of the 5400rpm drive is much greater than the 7200rpm drive. The increased areal density can more than make up for the decreased angular frequency.
For example, my Maxtor 5120 (5GB/platter if I remember correctly and 7200rpm) is much slower than my WD 45 (15GB/platter and 5400rpm) when opening large sound files in Sound Forge.
The 7200rpm drive will have an advantage in seek time though.
One way is if you are comparing maximum transfer rates and the areal density of the 5400rpm drive is much greater than the 7200rpm drive. The increased areal density can more than make up for the decreased angular frequency.
For example, my Maxtor 5120 (5GB/platter if I remember correctly and 7200rpm) is much slower than my WD 45 (15GB/platter and 5400rpm) when opening large sound files in Sound Forge.
The 7200rpm drive will have an advantage in seek time though.
The 5400RPM can have a faster track seek becuase finding a track is unrelated to spindle speed. The speed of the arm/heads moving could be higher on a 5400RPM than a 7200RPM, I think that's just theory though...because when you do a seek you have to factor in rotational latency which is a function of (primarily) spindle speed.
The only practical answer is hulk's. For example an IBM 40GV has 20gigs/platter, and lots of older 7200rpms are 5Gig or less (like his maxtor 5120). Because transfer rate is a function of spindle speed AND areal density (among other things) a lower RPM higher density drive can transfer faster.
The only practical answer is hulk's. For example an IBM 40GV has 20gigs/platter, and lots of older 7200rpms are 5Gig or less (like his maxtor 5120). Because transfer rate is a function of spindle speed AND areal density (among other things) a lower RPM higher density drive can transfer faster.
Rotational latency is completely controlled by spindle speed, it is considered to be 1/2 the time for 1 revolution. On average the required data will be 1/2 a rev away when the heads reach the correct cylinder. The difference between 5400 and 7200 is about 1.4 millisecs if I recall. (i'm too lazy to do the math). The RPMs has NOTHING to do with the drive seek time. The actual data transfer speed depends much more on areal density than the interface speed.
In other words, there is no reason to use a 5400rpm drive if you don't have to. Most new 7200rpm drives are 15-20GB/platter, no? However, if you're comparing a 5GB/platter 7200rpm drive to a 15GB/platter 5400rpm drive, of course the 5400rpm drive will have an advantage in certain areas.
Cretin
Cretin
Argo, what makes you think the seek times include latency ?? It certainly never used to include it. Total I/O time was always specd at Seek Time + Latency + Data Transfer. There is also a hidden component of Seek Time called, Head Settle Time, which is more important on writes than reads. Its possible that the Seek Time recorded by HDD test programs include latency but the drive nmanufacturers specs usually don't.
Cretin:
Tbere is NO performance advantage to the 5400 RPM spindle speed in of itself. Only when areal density is consider can some newer 5400 drives beat the less dense older 7200 RPM units. However, the practical consideration has been that the 7200 RPM rotation speed is often not worth the extra cost to manufacture the technology which translates into more $$/gig. 5400 RPM drives, depending on your needs may offer you the better bang for buck. (less $$ per gig)
But take a drive designed to run at 7200 RPM and if you could slow the motor down to 5400RPM without affecting the operation, it would seek slower than the 7200 RPM mode of operation. But almost never do storage companies create two clone models that vary only in RPM. THe drives are totally engineered for their rotation speed and areal density. So, bottom line is, go for the balance of performance and cost/gig that suits your needs.
Tbere is NO performance advantage to the 5400 RPM spindle speed in of itself. Only when areal density is consider can some newer 5400 drives beat the less dense older 7200 RPM units. However, the practical consideration has been that the 7200 RPM rotation speed is often not worth the extra cost to manufacture the technology which translates into more $$/gig. 5400 RPM drives, depending on your needs may offer you the better bang for buck. (less $$ per gig)
But take a drive designed to run at 7200 RPM and if you could slow the motor down to 5400RPM without affecting the operation, it would seek slower than the 7200 RPM mode of operation. But almost never do storage companies create two clone models that vary only in RPM. THe drives are totally engineered for their rotation speed and areal density. So, bottom line is, go for the balance of performance and cost/gig that suits your needs.
<< Argo, what makes you think the seek times include latency ?? >>
This is kind of a hazy definition..personally I consider seek time to be the time it takes to find a track, and after that you have rotational latency. Some people classify seek time as the time it takes to find a particular sector (byte? bit?) of the drive, in which case you have to take rotational latency into account on the seek time.
I've had many long debates with co-workers and "professionals", and students and professors at my old university on which is the correct definition, to little or no avail.
I have coined the terms "track" seek time, and "total" seek time to differientiate, but if someone knows the proper definition of seek time I'd love to know it.
<< Most new 7200rpm drives are 15-20GB/platter, no? >>
Of course. An IBM 75GXP is 15Gig/platter, and the WD Caviar BB series is 20Gig/platter.
There is no way a 15Gig/platter 9ms seek 7200rpm is going to be beaten by a 15Gig/platter 9ms seek 5400rpm. Depending on exactly what you are doing they might be close, or even the same, but the 5400 won't win.
He just asked if a 5400 can be faster, and there are circumstances where a 5400 *can* be...
Angrymarshmello
Lifer
So all in all they're about the same performance wise?
kami
Lifer
Platter size makes a huge difference.
I have two drives...a Maxtor DiamondMax Plus 45 which is 15gigs/platter and a IBM 22GXP which is I don't know how many per platter but it would be low..close to 5 i think. Both are 7200rpm and both running at ATA/66.
With the maxtor in HD Tach I get a maximum of 35megs/sec transfer and a minimum of 31megs/sec. That is pretty consistent if you ask me. With the IBM I get a maximum of 22megs/sec and a minimum of about 13megs/sec.
Both have equal burst rates... 60meg/sec. The maxtor is 2ms quicker in the access time. I bet the best 5400rpm drive out there (with a big platter) could match or even beat the 22GXP in certain areas.
I have two drives...a Maxtor DiamondMax Plus 45 which is 15gigs/platter and a IBM 22GXP which is I don't know how many per platter but it would be low..close to 5 i think. Both are 7200rpm and both running at ATA/66.
With the maxtor in HD Tach I get a maximum of 35megs/sec transfer and a minimum of 31megs/sec. That is pretty consistent if you ask me. With the IBM I get a maximum of 22megs/sec and a minimum of about 13megs/sec.
Both have equal burst rates... 60meg/sec. The maxtor is 2ms quicker in the access time. I bet the best 5400rpm drive out there (with a big platter) could match or even beat the 22GXP in certain areas.
Noriaki,
Can't claim to be an expert, but I have worked with disk drives since 1968. At that time they were as big as a washing machine, held 7 meg and used a hydraulic actuator !!. I have specialized in Disk Performance with both IBM and later with AMDAHL. Published Seek Times never included latency, Head Settle Time was frequently omitted by competing manufactures but it is important on writes. If the head is still 'shaking' after the seek, its not such a problem on reads, but on writes, the data may not be able to be read.
Can't claim to be an expert, but I have worked with disk drives since 1968. At that time they were as big as a washing machine, held 7 meg and used a hydraulic actuator !!. I have specialized in Disk Performance with both IBM and later with AMDAHL. Published Seek Times never included latency, Head Settle Time was frequently omitted by competing manufactures but it is important on writes. If the head is still 'shaking' after the seek, its not such a problem on reads, but on writes, the data may not be able to be read.