Samsung 830 or Crucial m4 256GB, any real world power consumption difference?

[bailybilko]

Im new to ssd, just installed a 830 128GB in one "movable" that's almost always plugged in.

So impressed by the difference i need an ssd for a tablet pc.

I was looking at either the Samsung 830 or Crucial m4.

I was reading some reviews that say the 830 has lower idle consumption, but the m4 is better in reading\writing.

Anyone know if this is a difference you would notice?

I looked at this chart, and the 830 has the lowest idle numbers, but highest read\write
http://www.anandtech.com/show/5719/ocz-vertex-4-review-256gb-512gb/9

 

[jwilliams4200]

In addition to idle power, what you need to look at is not really the power usage during read and write, but rather the bytes read or written for each unit of energy consumed (power is energy per time). MB/s per Watt (speed per power usage) is equivalent to bytes per energy, and is sometimes called energy efficiency.

If you compare the energy efficiency of the Samsung 830 and the m4, they are similar. The m4 is certainly NOT better than the Samsung.

http://www.behardware.com/articles/860-23/ssd-2012-roundup-sixteen-120-and-128-gb-sata-6g-ssds.html

 

[hhhd1]

for ssd, idle power usage is all that matters, because the drive stays idle most of the time.

some reviews suggest that the samsung 830 uses less power at idle than the crucial m4, samsung 830 is pretty much the lowest possible power usage at idle.

do some googling.

 

[jwilliams4200]

for ssd, idle power usage is all that matters, because the drive stays idle most of the time.

Not necessarily true. Since the power usage during writes is 5 to 10x as much as during idle, if your SSD is writing 10% of the time, then it will be using about double the power it would be if it were idling all the time.

Now, some people's SSDs may be writing less than 1% of the time, in which case the power usage during writes would be unimportant, as you say.

But whether that is true or not obviously depends on the workload that the SSD is subjected to.

 

[bailybilko]

Well, after googling the the correct phrases i found someone that asked the same on another forum also..

Ordered a 830 when it was in stock for 5 minutes since it was cheaper, hope it wasnt a shop error.

 

[hhhd1]

Not necessarily true. Since the power usage during writes is 5 to 10x as much as during idle, if your SSD is writing 10% of the time, then it will be using about double the power it would be if it were idling all the time.

Now, some people's SSDs may be writing less than 1% of the time, in which case the power usage during writes would be unimportant, as you say.

But whether that is true or not obviously depends on the workload that the SSD is subjected to.

10% ? the average usage is less than 0.1%

for an ssd with read/write speed = 500/175:
10% = constant read at 50mb/s ,
even the OS loading doesn't do that.

Loading xp uses an average of 5~6 mb/s for like 30 seconds (using vmware to see the read speed)

the ssd doesn't write allot, unless you are downloading a file with huge speed, or doing a huge setup. and even that doesn't usually last for long.

so it is only about the reading and idling.



bottom line:
The disk idles 99.9% of the time due to its very high read speed.
The exceptions:
- you are constantly loading and setting up OS'es (1~8% usage)
- you are constantly loading games, and not just playing (~10% usage, depends on the processor)
- you have very low ram, and using a large page file, while doing ram intensive applications.
- doing file transfer to another hard disk (10~20% usage)
- doing file transfer to another SSD (50~100% usage)

 

[hhhd1]

the crucial m4 128gb:
assuming a constant write at 10%:

write speed = 175
constant writing at 10% = 17.5mb
writing in 1 min = 60 * 17.5 = 1050mb
writing in 1 day = 86,400 * 17.5 = 1,512,000mb

assuming a write implication of 4:
daily actual write = 6,048,000mb

the crucial m4 128gb (==122,000mb) has 3000 life cycles,
means that the ssd is capable of writing 122000*3000 = 366,000,000 to reach its rated life

366,000,000 / 6,048,000 = 60

all this means:
constant writing at 17.5 mb, or at 10% of available write speed, will make it reach its maximum expected life in 60 days, (theoretically KiLL the ssd in about 60 days).

99.9+% of the users will not reach those limits, because 99.9+% of the users do not write that amount of data.

 

[jwilliams4200]

99.9+% of the users will not reach those limits, because 99.9+% of the users do not write that amount of data.

Do you list "making up BS statistics" as one of your skills on your resume?

It is not hard to think of workloads that will write 10% or more of the time the SSD is in use (which need not be 24/7 which you assume for some strange reason) -- VM disk with a lot of swapping, live TV video buffer, heavily used database server, photoshop scratch disk, backup drive, video editing, etc.

Just because you do not use that particular workload does not mean that no one does. I would be willing to posit that the majority of people probably do not have such an SSD intensive workload, but I certainly will not make up BS statistics. Besides, even if your made-up statistic were true, there would still be some people with workloads for which idle power is NOT all that matters, which was the point that I made before you went off on an unrelated tangent, ranting and raving to no useful purpose.

 

[StarTech]

Do you list "making up BS statistics" as one of your skills on your resume?

It is not hard to think of workloads that will write 10% or more of the time the SSD is in use (which need not be 24/7 which you assume for some strange reason) -- VM disk with a lot of swapping, live TV video buffer, heavily used database server, photoshop scratch disk, backup drive, video editing, etc.

Just because you do not use that particular workload does not mean that no one does. I would be willing to posit that the majority of people probably do not have such an SSD intensive workload, but I certainly will not make up BS statistics. Besides, even if your made-up statistic were true, there would still be some people with workloads for which idle power is NOT all that matters, which was the point that I made before you went off on an unrelated tangent, ranting and raving to no useful purpose.

I don't know it is 99.99% or 97.35%. I do not care. I am of the opinion that hhhd1 is more on target than you. Making up or not statistics, a 10% busy ratio with SSD speeds is an extreme, very high hit rate. More so for a tablet which is the OP's interest. I can see very rare, isolated situations where someone chooses to burn an SDD for some application, just because of the speed benefit/low cost, but most users have to work rather hard to push a sustained 10% utilization rate.

 

[jwilliams4200]

but most users have to work rather hard to push a sustained 10% utilization rate.

So you agree with me, since you did not say "all users". If you disagree, please quote the exact statement that you disagree with.

 

[Soulkeeper]

they are similar enough to where i'd purchase on no other criteria than price.

 

[Kambz]

Yeah both seem really similar to me. I have crucial tho :O

 

[StarTech]

So you agree with me, since you did not say "all users". If you disagree, please quote the exact statement that you disagree with.

I got from your posts that you think the power consumption due to drive activity is significant, and you do so using a sample 10% rate that is not normal at all. I think it is unrealistic, and if the 99.99% may be wrong, so is the 10%. The OP asked about power use in a tablet. I cannot see a tablet putting 10% load on the SSD and the numbers thrown in the other post are trying to say that a drive with that activity would wear down very soon, which we don't see reported normally.

So I expressed my opinion. You can like it or not. If I happen to be wrong, it is not a big deal; I will not go to hell because of it. I do not come to the forums to fight, and I do not take it as if my intelligence was at stake, not I think you are less intelligent because I think you may be wrong on this point. I have been wrong many times, and have not exhausted my quota yet. You will notice by my profile, that I have been an AT member since the late 90's and that I have less than 1,000 posts. I just don't keep going back and forth because it does not fill my life.

 

[jwilliams4200]

I got from your posts that you think the power consumption due to drive activity is significant, and you do so using a sample 10% rate that is not normal at all.

You really need to stop misquoting me. I specifically asked you to quote me where you disagreed, but instead you make up words that I did not write. Kindly stop doing that.