Show us your SSD performance 2

[essenbe]

jonnyhillow, That's not too bad, don't worry much about it, it's still 10 times faster than a mechanical hard drive.

 

[jonnyhillow]

jonnyhillow, That's not too bad, don't worry much about it, it's still 10 times faster than a mechanical hard drive.

You sure about that , maybe i'm wrong but comparing the below to mine is depressing to say the least .

I mean i'm not even remotely close , man this sucks.

 

[HarriePateman]

If you guys are worried yours is slow, do you have rapid mode enabled?

There is a tutorial on optimizing your ssd but I'm at work so I will post the link later!

 

[paulpicks21]

jonnyhillow, That's not too bad, don't worry much about it, it's still 10 times faster than a mechanical hard drive.

You sure about that , maybe i'm wrong but comparing the below to mine is depressing to say the least .

I mean i'm not even remotely close , man this sucks.

Jonny, your score is completely fine and normal.

Steve's score and many of the recent ones on here are not normal. Steve is running Raid0 on 2 Samsung drives with something called "Rapid mode" enabled, this uses the Ram as a kind of cache disk. The numbers are massively inflated due to this. The same with most of the recent high scores, we all have Samsung drives with Rapid mode enabled. My drive for example will score around 900ish with Rapid mode disabled.

Harry, looking at his screetshot he has an Intel drive. Rapid mode comes with Samsung SSD's only.

 

[jonnyhillow]

I have a question, as you can see i always have the same amount of disk space taken up over the years , it never grows and always stays the same . Even on my standard non SSD drive i always had tons and tons of space available.

This being the case would a 128 gb SSD be appropriate or should i always opt for a 240 or larger, any performance differences.

I see a Samsung 128 EVO for $69.00 online.

 

[HarriePateman]

jonnyhillow, That's not too bad, don't worry much about it, it's still 10 times faster than a mechanical hard drive.

You sure about that , maybe i'm wrong but comparing the below to mine is depressing to say the least .

I mean i'm not even remotely close , man this sucks.

Jonny, your score is completely fine and normal.

Steve's score and many of the recent ones on here are not normal. Steve is running Raid0 on 2 Samsung drives with something called "Rapid mode" enabled, this uses the Ram as a kind of cache disk. The numbers are massively inflated due to this. The same with most of the recent high scores, we all have Samsung drives with Rapid mode enabled. My drive for example will score around 900ish with Rapid mode disabled.

Harry, looking at his screetshot he has an Intel drive. Rapid mode comes with Samsung SSD's only.

Ops sorry, i was at work and i couldn't see the image

 

[ZaLiTH]

This being the case would a 128 gb SSD be appropriate or should i always opt for a 240 or larger, any performance differences.

Not sure if the same logic applies to all SSDs, but with Crucial's M500 range, the only performance difference between the different sizes is the faster write speeds. I suspect this may have something to do with the overall density of the data storage, but that's just my best guess..

 

[essenbe]

Zalith, sometimes larger drives get better performance. Usually a 256 will give you higher scores. But, never confuse the benchmark scores to actual usage performance. Sure, those big scores are nice, but totally irrelevant to actual usage. An OS drive runs 90%+ of the time on the 4kb read speed and response times. That's all that matters. I showed 2 benchmarks, the rapid mode and a raid 0 mode. The rapid mode is totally useless as the test only tests the ram cache, which is way faster than any SSD. The sata ports cannot even come close to those speeds. So, the rapid scores mean nothing. I also showed a raid0 score around 1800-1900. Those scores are meaningful because it is 2 sata ports and the 1000MB/s it shows is most likely close to accurate for 2 sata ports. But, even in the raid 0 scores, look at the 4kb read speed and the access times. They are no higher than a single drive. So, for 90% of the time as an OS drive, raid is no faster and much more dangerous. Just remember that no matter what scores you see the 4kb read speed and the access times are the only scores that matter. The other scores would only matter if you were transferring 25 GB files all day, and you would still be limited by the write speed of the drive you are transferring to.

 

[ZaLiTH]

Zalith, sometimes larger drives get better performance.
[snip]

Thanks for the clear explanation.. I am still learning the finer details of the SSD world, but it is definitely making more and more sense as I go.

One thing I have worked out beyond a shadow of a doubt; many of the drives out there obliterate my M500 in the performance department...

As you pointed out though, there are many limiting factors. Since I have no plans to RAID any of them, their max 4k sequential read speed of 500MB/s is more than good enough for me.

I'll have a few interesting benchies to post in the coming weeks.. Moving house this weekend, then next weekend I'll (hopefully) get to build my new beast. Can't wait to see my games loading from an SSD as well.

 

[essenbe]

Zalith, sometimes larger drives get better performance.
[snip]

Thanks for the clear explanation.. I am still learning the finer details of the SSD world, but it is definitely making more and more sense as I go.

One thing I have worked out beyond a shadow of a doubt; many of the drives out there obliterate my M500 in the performance department...

As you pointed out though, there are many limiting factors. Since I have no plans to RAID any of them, their max 4k sequential read speed of 500MB/s is more than good enough for me.

I'll have a few interesting benchies to post in the coming weeks.. Moving house this weekend, then next weekend I'll (hopefully) get to build my new beast. Can't wait to see my games loading from an SSD as well.

The max actual transfer speed of a sata III port is around 475-515 MB/s, depending on your motherboard and several other things, but that is around the max even the fastest SSD can transfer data. The benchmark scores are really nothing but show. What an OS SSD does is the 4kb read speeds and access times. No matter if my Raid score is way faster than your score, but look at only the 4kb read speed and the access times. There is only slight differences. I will bet you if I took my Raid set up and put it in your machine and didn't tell you, you would turn your computer on the next day and notice no difference in actual usage. That is how much difference benchmark scores matter. So, don't be discouraged because yours benchmark lower than others. It does not matter at all in actual usage. I have 2 or 3 Crucial drives, in other machines. This Samsung 840 Pro will eat them alive in any benchmark you want them to run. In actual usage, there is little to no difference. Sometimes I can tell a slight difference, but 95% of the time there is none. I don't like raid and don't like running it and advise everyone else not to. The only reason I did it is I have 2 X 256GB SSDs. My OS takes up about 40-45GB which leaves a lot of wasred space, but my Games or data will not fit on the other 256 GB drive or on the extra space of my OS drive. So, basically I have a whole lot of expensive space left over and pretty much useless, for the way I want to organize things. So, I raided them to give me close to about 480-485GB. With that space, I can run my OS and games on the 2 SSDs. But, I am taking a big risk, but I backup every night and only have the OS and games on the raid. So, if one died today, it would cost me a clean install which I have down to a fine science. I could be up and functional in an hour after the updates are finished, and most of those I have in the ISO I use for installs. Just so you understand. Benchmarks mean nothing but bragging rites. There are only 2 numbers that really matter.

 

[ZaLiTH]

I will bet you if I took my Raid set up and put it in your machine and didn't tell you, you would turn your computer on the next day and notice no difference in actual usage. That is how much difference benchmark scores matter.

I'm definitely not discouraged, don't worry.. As impressive as many of these benchmarks are, they're much like the WEI story; it's mostly marketing gimmicks at the end of the day.

With that space, I can run my OS and games on the 2 SSDs. But, I am taking a big risk, but I backup every night and only have the OS and games on the raid. So, if one died today, it would cost me a clean install which I have down to a fine science.

My current machine's Win folder is sitting at about 27GB at the moment, which is pretty average for one of my setups, so my 120GB M500 (with desktop apps) is only half full. My games fit quite comfortably on my current 500GB 'Cuda, so I won't have any issues switching over to a 480GB M500.

Years back in WinXP's glory days, my brother had a pretty sweet backup/restore procedure.. In a nutshell, he had a couple scripts saved that would update the paths for his program, profile and data folders, which were stored on the D:/. If he ever had to format/reinstall, he would simply wipe the primary drive, reinstall, run his scripts and restart. The only thing we never quite got around to doing was automating the whole process. Fun times..

 

[jonnyhillow]

Zalith, sometimes larger drives get better performance.
[snip]

Thanks for the clear explanation.. I am still learning the finer details of the SSD world, but it is definitely making more and more sense as I go.

One thing I have worked out beyond a shadow of a doubt; many of the drives out there obliterate my M500 in the performance department...

As you pointed out though, there are many limiting factors. Since I have no plans to RAID any of them, their max 4k sequential read speed of 500MB/s is more than good enough for me.

I'll have a few interesting benchies to post in the coming weeks.. Moving house this weekend, then next weekend I'll (hopefully) get to build my new beast. Can't wait to see my games loading from an SSD as well.

The max actual transfer speed of a sata III port is around 475-515 MB/s, depending on your motherboard and several other things, but that is around the max even the fastest SSD can transfer data. The benchmark scores are really nothing but show. What an OS SSD does is the 4kb read speeds and access times. No matter if my Raid score is way faster than your score, but look at only the 4kb read speed and the access times. There is only slight differences. I will bet you if I took my Raid set up and put it in your machine and didn't tell you, you would turn your computer on the next day and notice no difference in actual usage. That is how much difference benchmark scores matter. So, don't be discouraged because yours benchmark lower than others. It does not matter at all in actual usage. I have 2 or 3 Crucial drives, in other machines. This Samsung 840 Pro will eat them alive in any benchmark you want them to run. In actual usage, there is little to no difference. Sometimes I can tell a slight difference, but 95% of the time there is none. I don't like raid and don't like running it and advise everyone else not to. The only reason I did it is I have 2 X 256GB SSDs. My OS takes up about 40-45GB which leaves a lot of wasred space, but my Games or data will not fit on the other 256 GB drive or on the extra space of my OS drive. So, basically I have a whole lot of expensive space left over and pretty much useless, for the way I want to organize things. So, I raided them to give me close to about 480-485GB. With that space, I can run my OS and games on the 2 SSDs. But, I am taking a big risk, but I backup every night and only have the OS and games on the raid. So, if one died today, it would cost me a clean install which I have down to a fine science. I could be up and functional in an hour after the updates are finished, and most of those I have in the ISO I use for installs. Just so you understand. Benchmarks mean nothing but bragging rites. There are only 2 numbers that really matter.

Makes sense but doesn't explain the congrats and high fiving for the people who post the most impressive benchmarks .

Must just be a " you think thats good , check mine out " thing . Unfortunately for me, my drive also comes in dead last with the two scores that matter , damn this sucks.

 

[essenbe]

Let me show you how bad 'it sucks' Here is a Western Digital 1 TB WD1002FAEX sata III drive. Supposedly one of the faster desktop drives at the time. You tell me which you would rather have, this or yours.



I suspect you will say, mine's not too bad at all. The Benchmarks are just for bragging rights. They mean nothing and most of the people who run them (me included) are just trying to show off. Knowledgeable people who are running them know, they have no bearing on real world performance. And a few kb/s make no difference in the real world. The worst SSD is 10 times better than the best mechanical hard drive.

 

[kbrady1979]

Steve is 100% correct. I was going to type up a post but it seems he's said everything almost word for word that I was going to say. He even posted a benchmark of a regular HDD, which I was going to do as well! Jonny, those scores are fine and indicate no problem with the drive in my opinion. Benchmarks sometimes show us hidden issues with hardware, which is what they are really useful for to be honest. The only thing I will point out is you are using the generic Microsoft SATA controller driver(msahci in the top left corner) and we have found that USUALLY the newer Intel Rapid Storage Technology driver works better. I say usually because sometimes it makes things worse.....most of the time it works great. I would at least look into it.

 

[jonnyhillow]

Let me show you how bad 'it sucks' Here is a Western Digital 1 TB WD1002FAEX sata III drive. Supposedly one of the faster desktop drives at the time. You tell me which you would rather have, this or yours.

View attachment 334387

I suspect you will say, mine's not too bad at all. The Benchmarks are just for bragging rights. They mean nothing and most of the people who run them (me included) are just trying to show off. Knowledgeable people who are running them know, they have no bearing on real world performance. And a few kb/s make no difference in the real world. The worst SSD is 10 times better than the best mechanical hard drive.

After thinking about it some more maybe i rushed to judgment , i guess mine is just fine after all lol.

 

[jonnyhillow]

Steve is 100% correct. I was going to type up a post but it seems he's said everything almost word for word that I was going to say. He even posted a benchmark of a regular HDD, which I was going to do as well! Jonny, those scores are fine and indicate no problem with the drive in my opinion. Benchmarks sometimes show us hidden issues with hardware, which is what they are really useful for to be honest. The only thing I will point out is you are using the generic Microsoft SATA controller driver(msahci in the top left corner) and we have found that USUALLY the newer Intel Rapid Storage Technology driver works better. I say usually because sometimes it makes things worse.....most of the time it works great. I would at least look into it.

Thank you, i appreciate it.

 

[AntiPCW]

Samsung 840 PRO 256Gb

Great test. Here's my Samsung 840PRO 256 score.

 

[A Guy]

Posting just to make jonnyhillow feel better Intel X25M Gen2 80GB



A Guy

 

[kbrady1979]

Funny thing about that Bill, is that you won't notice a nickels worth of difference between that and my Samsung 830! I have an Intel X-25M 120GB and besides higher scores in benchmarks, there is no noticeable difference.

 

[jonnyhillow]

Posting just to make jonnyhillow feel better Intel X25M Gen2 80GB

View attachment 334437

A Guy

I have to be honest , i would be outraged and ready to kill if i got those scores , thank God some of you had good explanations .

My biggest issue is it really doesn't feel faster than my mechanical drive , i know it is but i wish it was more noticeable.