Can i trust my Samsung Magician performance test?

[iYCloud]

I have done several benchmarks using Samsung Magician to see speed of my Samsung 840 EVO SSD. I have enable RAPID Mode. Benchmark Results are amazing.

 

Read / Write (MB/s)

4094 / 2833

4036 / 3192

3540 / 2645

3776 / 2901

3371 / 2511

 

 

[WhatTheQuack]

Because you have RAPID mode enabled, you are basically running the test on your RAM (RAPID mode caches RAM with your SSD) so these results are accurate. However this does exceed the maximum speed of SATA 3 (I think) so you probably won't notice much of a difference if you turned RAPID mode off. That way you will be running the test from the SSD itself.

 

Also welcome to the forum!

[ericlee30]

 However this does exceed the maximum speed of SATA 3 (I think) so you probably won't notice much of a difference if you turned RAPID mode off.

 

 

You are correct sir

[Enderman]

However this does exceed the maximum speed of SATA 3 (I think) so you probably won't notice much of a difference if you turned RAPID mode off.

Could you explain what you mean by this?

There is a big difference between 500MB/s and 4000MB/s

[ericlee30]

Could you explain what you mean by this?

There is a big difference between 500MB/s and 4000MB/s

SATA 3 has a cap on it, if you go over that cap, then you will not get any better results then you already have. If that make since.

[WhatTheQuack]

Could you explain what you mean by this?

There is a big difference between 500MB/s and 4000MB/s

SATA 3 has a maximum theoretical transfer speed of 768MB/s, so having your SSD perform faster than this won't make a difference if it is bottle-necked by SATA anyway. However where you will notice the speed difference is from files that are stored in the RAM cache. However this will most likely be smaller files because the cache size is relatively small itself. So the difference here will be, for example, taking 0.1 seconds to load a file, or taking 0.0001 seconds, which is pretty much negligible anyway. 

[Enderman]

SATA 3 has a cap on it, if you go over that cap, then you will not get any better results then you already have. If that make since.

But the 4000MB/s is coming from the ram, not the SSD. It is not bottlenecked by sata 3

What RAPID does is it stores frequently used data on a ramdisk instead of the SSD so that when that data is needed it can be accessed at 4GB/s+

Obviously you cna see from the read/write tests that is IS a huge improvement over sata3, because it doesnt use sata 3.

[Enderman]

SATA 3 has a maximum theoretical transfer speed of 768MB/s, so having your SSD perform faster than this won't make a difference if it is bottle-necked by SATA anyway. However where you will notice the speed difference is from files that are stored in the RAM cache. However this will most likely be smaller files because the cache size is relatively small itself. So the difference will be taking 0.1 seconds to load a file, or taking 0.0001 seconds, which is pretty much negligible anyway. 

By turning RAPID off you only use the SSD, but with RAPID on you get ramdisk speeds because it doesnt use sata 3

the ram disk cache size is also proportional to the amount of ram in your system, so pretty big files can be stored there.

There is a big difference between a .1 and a.0001 second load time. Same as the is between a 1000 second and 1 second load time.

[WhatTheQuack]

By turning RAPID off you only use the SSD, but with RAPID on you get ramdisk speeds because it doesnt use sata 3

the ram disk cache size is also proportional to the amount of ram in your system, so pretty big files can be stored there.

There is a big difference between a .1 and a.0001 second load time. Same as the is between a 1000 second and 1 second load time.

While it is a big difference, it is not nearly as noticeable than a 1000 second vs 1 second load time, because anything around or below 0.1 seconds will feel instant anyway

(I know it doesn't use SATA 3)

[Enderman]

While it is a big difference, it is not nearly as noticeable than a 1000 second vs 1 second load time, because anything around or below 0.1 seconds will feel instant anyway

(I know it doesn't use SATA 3)

Linus speaks from experience that it is VERY noticeable to have 8 SSDs in raid 0 vs a single ssd. And he still gets less than 4GB/s

If anything below .1 seconds felt instant we wouldn't need anything over 10fps. Absolutely not true.

[ViruZ_HUN]

SATA 3 has a maximum theoretical transfer speed of 768MB/s, so having your SSD perform faster than this won't make a difference if it is bottle-necked by SATA anyway. However where you will notice the speed difference is from files that are stored in the RAM cache. However this will most likely be smaller files because the cache size is relatively small itself. So the difference here will be, for example, taking 0.1 seconds to load a file, or taking 0.0001 seconds, which is pretty much negligible anyway. 

Well There is already Sata Express. Hope that Sata express will bring some SSDs soon

[Unhelpful]

Could you explain what you mean by this?

There is a big difference between 500MB/s and 4000MB/s

 

Linus speaks from experience that it is VERY noticeable to have 8 SSDs in raid 0 vs a single ssd. And he still gets less than 4GB/s

If anything below .1 seconds felt instant we wouldn't need anything over 10fps. Absolutely not true.

It still goes past the maximum speed of the interface.

 

Linus has a PCI-E RAID Card.

 

Although for the most part I agree with you.

[ericlee30]

But the 4000MB/s is coming from the ram, not the SSD. It is not bottlenecked by sata 3

What RAPID does is it stores frequently used data on a ramdisk instead of the SSD so that when that data is needed it can be accessed at 4GB/s+

Obviously you cna see from the read/write tests that is IS a huge improvement over sata3, because it doesnt use sata 3.

I see your point, I dont nitice any diffrance in my Evo, when in RAPID mode but then again for some odd reason without my EVO being in RAPID mode gets this.

 

 

Honestly, I have no clue as to why it reads so high not in Rapid mode.

[Enderman]

I see your point, I dont nitice any diffrance in my Evo, when in RAPID mode but then again for some odd reason without my EVO being in RAPID mode gets this.

 

 

 

Honestly, I have no clue as to why it reads so high not in Rapid mode.

That is messed up. Have you tried crystaldiskmark? See if it gives you the same results.

[WhatTheQuack]

Linus speaks from experience that it is VERY noticeable to have 8 SSDs in raid 0 vs a single ssd. And he still gets less than 4GB/s

If anything below .1 seconds felt instant we wouldn't need anything over 10fps. Absolutely not true.

You can't really compare those two things. Loading a small file which takes 0.1 second will feel instant. If you load small files over and over again, you will start to notice the difference between 0.1 second per file and 0.0001 second per file. With a game you will notice lower fps because there is a display of images over and over again. You are unlikely to load small files over an over so the overall feel of RAPID mode vs no RAPID mode will be difficult to notice to the average person (Linus is not average, he has used his RAID array for years and so has got used to the slightly faster 'load times', therefore he is more likely to notice a difference on a slower system.

[Enderman]

You can't really compare those two things. Loading a small file which takes 0.1 second will feel instant. If you load small files over and over again, you will start to notice the difference between 0.1 second per file and 0.0001 second per file. With a game you will notice lower fps because there is a display of images over and over again. You are unlikely to load small files over an over so the overall feel of RAPID mode vs no RAPID mode will be difficult to notice to the average person (Linus is not average, he has used his RAID array for years and so has got used to the slightly faster 'load times', therefore he is more likely to notice a difference on a slower system.

Exactly so if one uses RAPID mode they will notice a difference when switching it off.

What makes no sense is 'RAPID mode makes no difference in speed because of sata 3"

[ericlee30]

That is messed up. Have you tried crystaldiskmark? See if it gives you the same results.

Yep, trid it when i first started noticing it, with crystaldiskmark its around the same results.

[WhatTheQuack]

Exactly so if one uses RAPID mode they will notice a difference when switching it off.

What makes no sense is 'RAPID mode makes no difference in speed because of sata 3"

Alright, fair enough I'll admit it, I was wrong.

 

But why are we having this argument? The OP's question has been answered, I doubt he really cares about this...

[Enderman]

Yep, trid it when i first started noticing it, with crystaldiskmark its around the same results.

That is weird. It might be some 'boost' mode that samsung magician does, because there is no way you can get those speeds from sata 3. It might be doing a small ram cache even with RAPID off. IDK if there is an option to disable that though...cause if you cant then that is some cheaty way to fake the SSD read write speeds...

[Enderman]

Alright, fair enough I'll admit it, I was wrong.

 

But why are we having this argument? The OP's question has been answered, I doubt he really cares about this...

Well there seems to be some strange ram cache even with RAPID off that ericlee found. He gets over 1GB/s with RAPID off ene though that is obviously not possible over sata3...

Samsung is up to something here...

[WhatTheQuack]

Well there seems to be some strange ram cache even with RAPID off that ericlee found. He gets over 1GB/s with RAPID off ene though that is obviously not possible over sata3...

Samsung is up to something here...

Huh, that is strange. My 840 EVO has completely normal speeds of around 400-500MB/s without RAPID mode. I wonder If it wasn't disabled properly

[ericlee30]

That is weird. It might be some 'boost' mode that samsung magician does, because there is no way you can get those speeds from sata 3. It might be doing a small ram cache even with RAPID off. IDK if there is an option to disable that though...cause if you cant then that is some cheaty way to fake the SSD read write speeds...

Thats honestly what i was thinking too, but I cant find any proff or anything that could be helping the Evo, I do now something not right, just cant find what it is, I have even reformated and just installed just the Samsung magician software and crystaldiskmark and still get the same results with nothing else installed

[WhatTheQuack]

Thats honestly what i was thinking too, but I cant find any proff or anything that could be helping the Evo, I do now something not right, just cant find what it is, I have even reformated and just installed just the Samsung magician software and crystaldiskmark and still get the same results with nothing else installed

 

Well I guess if the numbers are true, It's not exactly a bad thing right?

[ericlee30]

Well I guess if the numbers are true, It's not exactly a bad thing right?

You are absulutaly right .