Samsung Rapid mode improves gaming performance?

[Shou Miko]

I been wondering about the above, that is if Samsung's Rapid mode really do give u any improvement in gaming performance at all?

Currently i am running a Intel SSD 530 480GB for gaming and wondering to get a 1TB just for being able to have more games installed on my SSD and than with the price of the EVO 850 being the best with performance and a solid drive would the Rapid mode be a boost or just waste of memory?

I have a EVO 840 250GB in my work laptop with Rapid mode enabled, but running normal desktop with all the programs and waterfox I can't say i feel it's faster then without.

 

[Toothless]

no. it does nothing for performance

 

[Bill_Bright]

It does help with performance - but that is with disk access performance. But once your program is loaded into RAM, disk access is rare. So this is not likely anything you will see - especially if you have a decent amount of RAM.

 

[sttubs]

I did not notice any difference using rapid mode. But you can't perform backup of drive with rapid mode enabled though, maybe a 3rd party software might.

 

[AsRock]

Nope, not a single bit. how ever the 850's are solid drives all the same.

 

[Shou Miko]

Nope, not a single bit. how ever the 850's are solid drives all the same.

Ik they r solid otherwise i wouldn't use them

but always good to get confirmed by a lot of ppl they r solid meaning that Samsung is doing their job of making good ssd's

 

[Bill_Bright]

I have used Samsung 840 Pros and Evos and my latest builds have 850 Pros and Evos and I think they are great. Now, I don't have identical computers with different brand SSDs to do a side by side comparison, but frankly I don't care. My slowest SSD blows the socks off my fastest 10K drive. Any difference is SSDs is nothing my mind can see - except on paper and in benchmarking programs and they are artificial anyway.

 

[Vayra86]

Once you've upgraded to any decent SSD that doesn't have a semi-broken controller or doesn't run over a semi-broken controller (OCZ and Marvell, I'm looking at you) any further performance gains will only show in server/workstation/heavy applications or specific workloads with extreme amounts of constant disk access, which are extremely rare in a regular consumer environment.

Games will load faster, generally, but streaming data in-game is always loaded from RAM so it's a non issue beyond loading a level or map.
Applications will start faster, but in-app performance is loaded from RAM, same deal

I could go on... but you get the point. SSD's only affect boot & loading times, and uncached items that are requested from disk.

 

[Shou Miko]

Once you've upgraded to any decent SSD that doesn't have a semi-broken controller or doesn't run over a semi-broken controller (OCZ and Marvell, I'm looking at you) any further performance gains will only show in server/workstation/heavy applications or specific workloads with extreme amounts of constant disk access, which are extremely rare in a regular consumer environment.

Games will load faster, generally, but streaming data in-game is always loaded from RAM so it's a non issue beyond loading a level or map.
Applications will start faster, but in-app performance is loaded from RAM, same deal

I could go on... but you get the point. SSD's only affect boot & loading times, and uncached items that are requested from disk.

I already have a SSD just for gaming, was just wondering about this topic game to mind.

 

[Bill_Bright]

I could go on... but you get the point. SSD's only affect boot & loading times, and uncached items that are requested from disk.

This is true - except, SSDs also consume less power, take up less space (thus less impact on desired air flow), generate less heat, don't make any noise and have longer life expectancies too. This longer life and lower energy costs off-set, to a great extent, their initial higher purchase costs.

But really, I think to say, "SSDs only affect boot and loading times, and uncached items that are requested from the disk" is really minimizing their advantage. Boot times generally are greatly reduced - something you will appreciate every time you sit down at your computer. Plus waking from sleep mode is even more pronounced. I note when I press a key to wake my computer, the greater delay is waiting for my monitors to finish waking - and they are pretty quick too.

Also, SSDs and page files are ideal for each other and regardless how much RAM you have installed, you should still have a page file. Those who say otherwise have never been able to produce a study, white paper, or review that recommends disabling the PF. It is just NOT true that performance improves by disabling the PF when you have lots of RAM installed.

So Windows always likes to see a Page File and after more than 20 years of fine tuning (since W95), Microsoft has it down pat (contrary to what some may want us to believe). Modern Windows will use the page file to temporarily stuff lessor priority "pages" of data into this portion of your "virtual memory" and does so very efficiently. For this reason, unless you really truly are a bona fide expert in virtual memory utilization, it is best to just let Windows manage it dynamically (dynamically because this is NOT a set and forget setting - contrary to what some may think too).

Also, the pages of data stuffed into the PF is "cached" data and cached data is "open" pages of data ready for immediate use. So having the PF on your SSD will also improve fetching that "cached" data compared to data cached on a slow hard drive based Page File. This is in addition to much faster fetching times for the other "uncached" data ("closed" files) Vayra86 mentioned. Note it takes extra time for "closed" files to be read into and "opened" into memory, and marked as open in the disks file tables.

Frankly, except for very tight budgets, I see no reason to use HDs anymore - except, maybe for mass storage of my digitized collection of my 600+ music CDs, and in my backup storage server (at least until I am ready to replace that too).

 

[RejZoR]

Doesn't RAPID mode use both, RAM and a single NAND cells to make such fast writes? Multi-cell NAND is treated as single cell. Losing some of the capacity in favor of speed because writing single cell is faster than fiddling with all 3 cells.

More info:
http://www.samsung.com/global/busin...ment/Samsung_SSD_Rapid_Mode_Whitepaper_EN.pdf

EDIT:
Ok, that's TurboWrite, though it works with RAPID mode together anyway.

 

[Vayra86]

This is true - except, SSDs also consume less power, take up less space (thus less impact on desired air flow), generate less heat, don't make any noise and have longer life expectancies too. This longer life and lower energy costs off-set, to a great extent, their initial higher purchase costs.

But really, I think to say, "SSDs only affect boot and loading times, and uncached items that are requested from the disk" is really minimizing their advantage. Boot times generally are greatly reduced - something you will appreciate every time you sit down at your computer. Plus waking from sleep mode is even more pronounced. I note when I press a key to wake my computer, the greater delay is waiting for my monitors to finish waking - and they are pretty quick too.

Also, SSDs and page files are ideal for each other and regardless how much RAM you have installed, you should still have a page file. Those who say otherwise have never been able to produce a study, white paper, or review that recommends disabling the PF. It is just NOT true that performance improves by disabling the PF when you have lots of RAM installed.

So Windows always likes to see a Page File and after more than 20 years of fine tuning (since W95), Microsoft has it down pat (contrary to what some may want us to believe). Modern Windows will use the page file to temporarily stuff lessor priority "pages" of data into this portion of your "virtual memory" and does so very efficiently. For this reason, unless you really truly are a bona fide expert in virtual memory utilization, it is best to just let Windows manage it dynamically (dynamically because this is NOT a set and forget setting - contrary to what some may think too).

Also, the pages of data stuffed into the PF is "cached" data and cached data is "open" pages of data ready for immediate use. So having the PF on your SSD will also improve fetching that "cached" data compared to data cached on a slow hard drive based Page File. This is in addition to much faster fetching times for the other "uncached" data ("closed" files) Vayra86 mentioned. Note it takes extra time for "closed" files to be read into and "opened" into memory, and marked as open in the disks file tables.

Frankly, except for very tight budgets, I see no reason to use HDs anymore - except, maybe for mass storage of my digitized collection of my 600+ music CDs, and in my backup storage server (at least until I am ready to replace that too).

I would be very interested in an in-depth testing to see how page file use impacts system responsiveness / application responsiveness with HDD versus SSD. I doubt there is much of a difference since, as you say, fetching already happens as much as possible without impacting user experience because that is how the page file works.

Either way the expected return on using RAPID mode for this purpose I think is non-existant.

 

[eidairaman1]

RAPID uses system ram, honestly I seen no difference in gaming performance, so Id say save the ram for other tasks than RAPID

 

[RejZoR]

Disk speed never affected game performance. Only games that may be affected are games where level data is streamed as you go and may be experienced as stuttering because HDD is struggling to keep up with data requests. For the rest, games preload all data into RAM anyway. So effectively, you only get longer level loads and nothing else.

 

[Jborg]

Your games will load faster, that's about it. A good example is loading a BF4 MP game, on an HDD it took about 35-40 seconds, on an SSD it takes about 7-10 seconds. I have 2 Samsung SSDs but have never messed around with the rapid mode.

 

[Bill_Bright]

I would be very interested in an in-depth testing to see how page file use impacts system responsiveness / application responsiveness with HDD versus SSD. I doubt there is much of a difference since, as you say, fetching already happens as much as possible without impacting user experience because that is how the page file works.

I don't see why you think there would be little difference. I bet there is a HUGE difference - especially with systems with low amounts of RAM. Of course it would depend on the scenario used for testing but to me, it only makes sense for scenarios that hit the PF often for performance to be better when using a SSD for your PF. In scenarios where little PF activity is used, then the difference would be marginal.

But the fact of the matter is, in either case, those would be artificial scenarios and since the vast majority of users use their computers for a variety of tasks, not sure any testing environment would be realistic. But regardless, I know of no one who does not accept that any disk activity is faster with SSDs (whether "noticeable" or not). Again, even the slowest SSD is significantly faster than the fastest 10K hybrid HDs in seek, access and data transfer times - plus they don't suffer from degraded performance due to fragmentation issues either.

One interesting read about SSDs being ideally suited for Page Files is, SSD FAQs, Should the pagefile be placed on SSDs? (scroll down to 10th FAQ).

Disk speed never affected game performance.

Never say never as absolutes are "always" () wrong! I agree that disk speed rarely plays a role in game performance but again, there are too many variables to make such a blanket statement. Not all games are the same, nor are all computers. There are still some computers out there being used for games that have tiny dual core CPUs and just 2GB of RAM with 5400RPM hard disks with only 8MB of on-board buffer. On some games, disk speed may cause lags and jerkiness.

 

[cdawall]

Makes no difference. If you want more performance get another solid state and use raid.

 

[Bill_Bright]

Your games will load faster, that's about it.

If talking ONLY about your game (assuming adequate CPU, RAM and graphics resources), then this is true. But an SSD offers many more advantages than just loading a game to say, "that's about it" or "makes no difference". If you don't believe that, then I think you (1) either have too little experience with a "full" SSD system or (2) you have been using only SSDs for so long, you forgot how slow hard drives affect MANY aspects of computing. In the case of (2), I recommend you replace your SSDs with HDs and be prepared to be disenchanted.

 

[cdawall]

If talking ONLY about your game (assuming adequate CPU, RAM and graphics resources), then this is true. But an SSD offers many more advantages than just loading a game to say, "that's about it" or "makes no difference". If you don't believe that, then I think you (1) either have too little experience with a "full" SSD system or (2) you have been using only SSDs for so long, you forgot how slow hard drives affect MANY aspects of computing. In the case of (2), I recommend you replace your SSDs with HDs and be prepared to be disenchanted.

I believe he is referring to gaming only. Which he is correct it will only affect load times, not FPS.

 

[qubit]

@puma99dk| I'll chip in and also say that it will only (potentially) improve load and save times, but not framerate.

 

[Bill_Bright]

I believe he is referring to gaming only. Which he is correct it will only affect load times, not FPS.

Thanks and I agree about FPS not being affected (again, assuming adequate CPU, RAM and graphics resources). But unless a person uses their computer ONLY for gaming and NOTHING else at all, then IMO, marginalizing the advantages of SSDs with such a narrow, single descriptor is doing SSDs a grave injustice.

There are a dozen or so pros for going with SSDs and AFAIK, only one con (for now) and that is $/GB. But those costs keep coming down rapidly.

 

[Jborg]

Thanks and I agree about FPS not being affected (again, assuming adequate CPU, RAM and graphics resources). But unless a person uses their computer ONLY for gaming and NOTHING else at all, then IMO, marginalizing the advantages of SSDs with such a narrow, single descriptor is doing SSDs a grave injustice.

There are a dozen or so pros for going with SSDs and AFAIK, only one con (for now) and that is $/GB. But those costs keep coming down rapidly.

Agreed completely. Was just responding directly to the thread title

 

[Vayra86]

I don't see why you think there would be little difference. I bet there is a HUGE difference - especially with systems with low amounts of RAM. Of course it would depend on the scenario used for testing but to me, it only makes sense for scenarios that hit the PF often for performance to be better when using a SSD for your PF. In scenarios where little PF activity is used, then the difference would be marginal.

But the fact of the matter is, in either case, those would be artificial scenarios and since the vast majority of users use their computers for a variety of tasks, not sure any testing environment would be realistic. But regardless, I know of no one who does not accept that any disk activity is faster with SSDs (whether "noticeable" or not). Again, even the slowest SSD is significantly faster than the fastest 10K hybrid HDs in seek, access and data transfer times - plus they don't suffer from degraded performance due to fragmentation issues either.

One interesting read about SSDs being ideally suited for Page Files is, SSD FAQs, Should the pagefile be placed on SSDs? (scroll down to 10th FAQ).

Never say never as absolutes are "always" () wrong! I agree that disk speed rarely plays a role in game performance but again, there are too many variables to make such a blanket statement. Not all games are the same, nor are all computers. There are still some computers out there being used for games that have tiny dual core CPUs and just 2GB of RAM with 5400RPM hard disks with only 8MB of on-board buffer. On some games, disk speed may cause lags and jerkiness.

You're right, on systems hungry for resources, the results may well be very different and no testing scenario really is realistic. But the point I was referring to was more in terms of this thread as well; on this forum people seem to have a tendency to over-think the effects of every setting and every new part that comes out. I understand this, because we all want to maximize performance, but very often the point is totally missed. As with SSD's, their use for gaming is limited and there is another thread up right now about NVME and why it won't take off - PC markets are in decline, saturated etc. The only reason for thát, is that the performance level we have reached now for the masses, is 'sufficient' and these little SSD debates kind of end there for most people. This goes for many other things as well, Windows is a good example. It works without any real interruptions, so people want it to keep working and stay the same.

*edit: this is also why we see all these new 'great revolutionary' technologies come out these days like VR, 3D, 4K, HDR and new REC specifications and all the little goodies that came to gaming the past few years. The market needs new toys to stay alive. I know a group of consumers is eager for something new but I think a lot of these things won't be large successes.

 

[trog100]

"Also, SSDs and page files are ideal for each other and regardless how much RAM you have installed, you should still have a page file. Those who say otherwise have never been able to produce a study, white paper, or review that recommends disabling the PF. It is just NOT true that performance improves by disabling the PF when you have lots of RAM installed."

hello again bill..

i havnt and aint going to write a white paper on windows and its swapfile.. and apart from saving disk space there isnt much to gain from turning it off.. this is for one simple reason.. given enough real ram windows never ever uses it.. he he..

for the last ten years or so i have run with the windows pagefile disabled.. my reasons are twofold.. first to prove my point many years ago and second to save disk space on my always small operating system drive or partition.. i keep it small for easy back ups..

so bill when you say windows uses and needs a swapfile you are quite simply wrong.. given enough ram windows does not need or use its swapfile.. and believe it or not it never does..

all you have to do is to simply disable your own and you will find out.. i currently have 32 gigs of ram but for quite a few years i ran win XP with no swap file and only 4 gigs of ram..

you are not on your own believing what you do that windows does need its swapfile.. much is written saying it does.. most of it is copy and paste mythology dating from the distant past..

one other thing.. windows 10 is very very protective as regards what it lets people do.. ask yourself a simple question.. if windows really does need an active swapfile.. why does it include the option to disable it..

my system is all solid state.. it has everything going for it except the price.. but all it really does in a practical sense is speed up boot and loading times.. it does speed up large game loading times a fair bit.. with other smaller files it dosnt make much difference..

i would like to have more ram.. enough to use a ram drive and pre-load a 40 gig game install directly into it.. i do have some software called dimm drive set up to do this but dont have enough ram.. you really need 64 gig.. my 32 gig is enough for smaller games but not the important ones..

i could ask you another simple question.. why would i say (given enough ram) windows dosnt need its swapfile unless i know for sure that what i am saying is true.. do you think i am too stupid to know.. he he he

no need to answer my last question bill.. you quite clearly do.. he he

trog

 

[OneMoar]

rapid mode is a glorified ram disk
it does exactly nothing