Samsung 840 vs Samsung 840 Pro

[GMPoisoN]

How large is the performance difference is there between the Samsung 840 and Samsung 840 Pro. I have a choice between a 128 GB 840 Pro or a 248 GB 840. With the 248 GB I could put Steam and a few other games on there, but if the Pro is that much faster, maybe I should go for that.

Thoughts?

Edit: May be able to spring for 248 GB Pro after downgrading GD-65 Gaming to GD-65.

 

[Soda-88]

The main factor between these two is the write endurance. 840 Pro uses MLC flash and 840 uses half as long lasting TLC flash. Now, taking into consideration you have double the capacity on non Pro SSD, both of them should be able to write equal absolute amount of data so I'd go with the 840.
Performance difference is pretty much irrelevant in a desktop environment. Just remember to treat your SSD well, i.e. no frapsing (near-raw video capturing) directly onto it.

 

[ronintexas]

The biggest advantage a SSD drive will give you is on the operating system, and commonly used programs. I have read many articles on reliability of the drives, failure rates, etc.... The current class of SSDs are almost as reliable (some argue more reliable) than a hard drive. The big issue is cost per GB. Current SSD drives are close to $1 per GB, standard hard drives are closer to $0.05 - $0.10 per GB. That makes the SSD approximately 10-20 times more expensive.

I use the 128GB Samsung 840 Pro, my wife has my old 128GB Samsung 840. The boot times as compared to a hard drive are astonishingly different, but when comparing SSD to SSD, there isn't a huge difference (30-45 seconds for HDD, 7-8 seconds for 840, 6-7 seconds for 840 Pro).

The biggest factor is going to be the reliability. The "torture" tests out there that simulate use are typically reading/writing to the drive at a pace that would take the average user 10+ years to accomplish what they do in days or weeks. Typically, the 840 Pro lasts twice as long as the 840...but a test lab isn't your living room. Since they have only recently (last 1-2 years) become common place, we are still at least 5-10 years from seeing true failure rates by consumer use.

I would recommend the 840 Pro 128GB, just a personal opinion. And just like AMD vs. Intel, or Ford vs. Chevy vs. Dodge, depending upon who you ask, you will get a different suggestion. LOL.

 

[GMPoisoN]

Thank you both, turns out after making a downgrade almost unnoticeable to another part, Im able to work in the 248 GB Pro. Both provided great information, only marking one of you as best answer over the other is because I don't want people relying to this solved post 😛

 

[RetiredChief]

Just a couple of things:
First I have 3 Crucial M4's 3 Samsung 830's and a pair of Samsung 840 Pros.
Using AS SSD the 4's and the Samsung 830 have a Total score of mid 700s, my 256 gig 840 Pro is low 1100s - BIG dif in bench mark; HOWEVER, you would be hard pressed to tell the diff in real life. Difference in Boot Time a couple of sec, for program loads my eyeballs are not callibrated down to 5 MilliSeonds - LOL.

2nd, My biggest priority in selecting a SSD is relability/Least user Problems. True reliability is not totally valid yet as none of the SATA III SSDs have been out for 5 Years.

I do NOT think that the "torture" Test are vaid, for two reasons.
1) Sample rate generally 1, Need multiple OFF-the-self samples not hand picked by manuf.
2) Probably a more valid reason is that the Torture test do NOT periodically stop and leave the SSD unpowered for a given time, then verify that the data is still vaild. To try to explain. You periodically power off and later power on. A NEW SSD will last for a considerable time (say 3 Months to a year) in an unpowered state. As the SSD nears it life cycle the amount of time in the unpowered state decreases considerably. And here-in lies the difference in testing. The continous read/writes whtout a power off an back on will go to X number of read/writes. In reality you may only get 1/2 to 3/4 of that number as when you power off, say at night, and power on the next day the cells have lost their charge.

This is very simular to a battery in your car nearing end of life. It's great you drive it, stop for gas and it restarts; However you get home turn car off and go to start it in the morning ony to find you have a dead battery.

In answer to your question, The 240 Gig 840 non pro should last for a considerale time say +5years (New tech by then would dictate buying the newest greatest anyway). And the size makes it a good choice. Myself, I would NOT buy the 840 non-pro, But then the diff in cost between the Pro and the Non pro is chump change for me, NOT true of everyone.

However; the Curcial M4 is everybit as good as the 840 Non-pro and their 256 gig M4 is often on sale for $170 which would be my choice (Samsung 830 would also be a good choice at the same price point - but is getting hard to find.

 

[GMPoisoN]

Just a couple of things:
First I have 3 Crucial M4's 3 Samsung 830's and a pair of Samsung 840 Pros.
Using AS SSD the 4's and the Samsung 830 have a Total score of mid 700s, my 256 gig 840 Pro is low 1100s - BIG dif in bench mark; HOWEVER, you would be hard pressed to tell the diff in real life. Difference in Boot Time a couple of sec, for program loads my eyeballs are not callibrated down to 5 MilliSeonds - LOL.

@nd, My biggest priority in selecting a SSD is relability/Least user Problems. True reliability is not totally valid yet as none of the SATA III SSDs have been out for 5 Years.

I do NOT think that the "torture" are vaid, for two reasons.
1) Sample rate generally 1, Need multiple OFF-the-self samples not hand picked by manuf.
2) Probably a more valid reason i that the Torture test do NOT periodically stop and leave the SSD unpowered for a given time, then verify that the data is still vaild. To try to explain. You periodically power off and later power on. A NEW SSD will last for a considerable time (say 3 Months to a year) in an unpoered state. As the SSD nears it life cycle the amount of time in the unpowered state decreases considerably. And here-in lies the difference in testing. The continous read/writes whtout a power off an back on will go to X number of read/writes. In reality you may only get 1/2 to 3/4 of that number as when you power off, say at night, and power on the next day the cells have lost their chare. This is very simular to a battery in your car nearing end of life. It's great you drive it, stop for gas and it restarts; However you get home turn car off and go to start it in the morning ony to find you have a dead battery.

In answer to your question, The 240 Gig 840 non pro should last for a considerale time say +5years (New tech by then would dictate buying the newest greatest anyway). And the size makes it a good choice. Myself, I would NOT, But then the diff in cost between the Pro and the Non pro is chump change for me, NOT true of everyone.

However; the Curcial M4 is everybit as good as the 840 Non-pro and their 256 gig M4 is often on sale for $170 which would be my choice (*30 would also be a good choice at the same price point - but is getting hard to find.

Thank you for that. Only going with the 256 Pro because the budget allow for it.

 

[MarkJohnson]

Just a couple of things:
First I have 3 Crucial M4's 3 Samsung 830's and a pair of Samsung 840 Pros.
Using AS SSD the 4's and the Samsung 830 have a Total score of mid 700s, my 256 gig 840 Pro is low 1100s - BIG dif in bench mark; HOWEVER, you would be hard pressed to tell the diff in real life. Difference in Boot Time a couple of sec, for program loads my eyeballs are not callibrated down to 5 MilliSeonds - LOL.

2nd, My biggest priority in selecting a SSD is relability/Least user Problems. True reliability is not totally valid yet as none of the SATA III SSDs have been out for 5 Years.

I do NOT think that the "torture" Test are vaid, for two reasons.
1) Sample rate generally 1, Need multiple OFF-the-self samples not hand picked by manuf.
2) Probably a more valid reason is that the Torture test do NOT periodically stop and leave the SSD unpowered for a given time, then verify that the data is still vaild. To try to explain. You periodically power off and later power on. A NEW SSD will last for a considerable time (say 3 Months to a year) in an unpowered state. As the SSD nears it life cycle the amount of time in the unpowered state decreases considerably. And here-in lies the difference in testing. The continous read/writes whtout a power off an back on will go to X number of read/writes. In reality you may only get 1/2 to 3/4 of that number as when you power off, say at night, and power on the next day the cells have lost their charge.

This is very simular to a battery in your car nearing end of life. It's great you drive it, stop for gas and it restarts; However you get home turn car off and go to start it in the morning ony to find you have a dead battery.

In answer to your question, The 240 Gig 840 non pro should last for a considerale time say +5years (New tech by then would dictate buying the newest greatest anyway). And the size makes it a good choice. Myself, I would NOT buy the 840 non-pro, But then the diff in cost between the Pro and the Non pro is chump change for me, NOT true of everyone.

However; the Curcial M4 is everybit as good as the 840 Non-pro and their 256 gig M4 is often on sale for $170 which would be my choice (Samsung 830 would also be a good choice at the same price point - but is getting hard to find.

Sorry to revive this thread, but I had an additional question.

My OCZ 240GB is full and I need a bigger SSD. I have noticed my OCZ vertex 3 has gotten slower. It is very laggy, even after reformat(secure erased). I assume this is from degradation over time that I keep reading about for all SSDs.

I guess my question is how well well are these new memories at keeping their speeds and, most importantly, latency over time.

I'm thinking of the pro, mostly for reselling when I need a bigger SSD. I assume it will hold it's value longer. But it's 50% more and not sure if it will hold it's value that well. If degradation is a big enough factor, then I'd gladly pay the extra 50%.