Samsung Magician 4.5 RAPID Mode – Evaluated With Demonstrated Speed Increase



For our last benchmark, we have decided to use PCMark 8 Extended Storage Workload in order to determine steady state throughput of the SSD.  This software is the longest in our battery of tests and takes just under 18 hours per SSD.  As this is a specialized component of PCMark 8 Professional, its final result is void of any colorful graphs or charts typical of the normal online results and deciphering the resulting excel file into an easily understood result takes several more hours.

There are 18 phases of testing throughout the entire run, 8 runs of the Degradation Phase, 5 runs of the Steady State Phase and 5 runs of the Recovery Phase.  In each phase, several performance tests are run of 10 different software programs; Adobe After Effects, Illustrator, InDesign, Photoshop Heavy and Photoshop Light, Microsoft Excel, PowerPoint and Word, as well as Battlefield 3 and World of Warcraft to cover the gaming element.

  • PRECONDITIONING -The entire SSD is filled twice sequentially with random data of a 128KB file size.  The second run accounts for overprovisioning that would have escaped the first;
  • DEGRADATION PHASE – The SSD is hit with random writes of between 4KB and 1MB for 10 minutes and then a single pass performance test is done of each application.  The cycle is repeated 8 times, and with each time, the duration of random writes increases by 5 minutes;
  • STEADY STATE PHASE – The drive is hit with random writes of between 4KB and 1MB for 45 minutes before each application is put through a performance test.  This process is repeated 5 times;
  • RECOVERY PHASE – The SSD is allowed to idle for 5 minutes before and between performance tests of all applications.  This is repeated 5 times which accounts for garbage collection; and
  • CLEANUP – The entire SSD is written with zero data at a write size of 128KB

In reading the results, the Degrade and Steady State phases represent heavy workload testing while the recovery phase represents typical consumer light workload testing.


As you can see, performance is recorded in terms of Bandwidth and Latency. Bandwidth (or throughput) represents the total throughput the drive is able to sustain during the tests during each phase. Latency, at least for the purposes of PCMark 8, takes on a different outlook and for this, we will term it ‘Total Storage Latency’.  Typically, latency has been addressed as the time it takes for a command to be executed, or rather, the time from when the last command completed to the time that the next command started.  This is shown below as ‘Average Latency’.

PCMark 8 provides a slightly different measurement, however, that we are terming as ‘Total Storage Latency’.  This is represented as being the period from the time the last command was completed, until the time it took to complete the next task; the difference of course being that the execution of that task is included in ‘Total Storage Latency’.  For both latency graphs, the same still exists where the lower the latency, the faster the responsiveness of the system will be.  While both latency charts look very similar, the scale puts into perspective how just a few milliseconds can increase the length of time to complete multiple workloads.

For a more in-depth look into Latency, Bandwidth, and IOPS check out our primer article on them here.


These results show the total average bandwidth across all tests in the 18 phases. In this graph the higher the result the better.

RAPID on vs off Total Bandwidth


These results show the average access time during the workloads across all tests in the 18 phases. In this graph the lower the result the better.

RAPID on vs off Average Latency Both


These results show the total access time across all tests in the 18 phases. In this graph the lower the result the better.

RAPID on vs off Total Latency

For this test we used the 120GB Samsung 850 EVO. At first, average bandwidth shows us two things. One, heavy workloads seem to result in lower average bandwidth with RAPID, and two, once the typical consumer light workloads begin, we can see it boosts performance to nearly 3x. Furthermore, we can see that the average latency without RAPID mode reaches up to almost .6ms in the Degrade and Steady State phases. Once RAPID is enabled, latency is cut down to just under .3ms. Over the Recovery phases we can see decent performance of around 0.072ms for latency without RAPID, however, once again, with it enabled we can see an average of 0.023ms for its latency performance. Finally, the total latency, results show that while bandwidth was less in steady state performance, overall, the reduction of average latency significantly speed up the workloads.


RAPID mode is a sweet extra feature that Samsung packed with their Magician Toolbox bundle. It allows for many Samsung SSD owners to enable RAM caching on their system in a quick and easy way. Once it is enabled, all the work is automated. The user can simply resume using the system as normal, but now performance is improved.

In testing, Samsung’s synthetic benchmark showed a night and day difference in performance with RAPID mode enabled, but it didn’t feel like it in our perceptible use of the system when enabled. To test out why, we benchmarked our startup time and application load times. As a result, startup was slower, but it was so minute a difference that it shouldn’t cause anyone to lose any sleep over it. Application load times also proved to gain no benefit, however, we were testing only the programs themselves at first. After testing some work files, it followed a similar trend. Recently accessed files should prove to have a faster retrieval speed after reboots, however they did not. Our testing methodology proved RAPID did not afford any benefit outside of Windows default caching behavior in these instances. It wasn’t until our workload testing was done that we saw a performance increase.

Our final testing with PCMark 8 gave us some better insight on how RAPID can improve a system’s performance. Average latency was cut in half and the total latency was cut down by more than that. The average bandwidth results proved something a bit different at first, steady state performance showed to be lower, but when in the recovery phase, average bandwidth shot up to speeds nearly triple to that of the system with RAPID disabled. Based upon our results, we can tell that RAPID actually does work and it really can significantly speed up real world performance.

So in all, RAPID does indeed increase a system’s performance. In day to day usage start up and opening application times will not be affected, however, when it comes to actually working on projects, RAPID will help to speed things up.

User Rating: 4.23 ( 8 votes)


  1. blank

    Took 1.5GB I guess out of my 8GB pc. RAPID is really nice but I think I can live without it too. All that matters is that one has to move from a HDD to SSD for atleast the boot drive. Have the 850 Pro myself. Team RED we are 🙂

  2. blank

    Hi Sean, what are your thoughts on RAID 0 Mode? I have 2 Samsung 120GB 850 EVOs. Should i leave them separate, first one being the system disk with RAPID enabled and second one for programs or should I RAID 0 them? I don’t care for the safety of the data, just speed. I do some prepress design and gaming. When in RAID 0 mode, Samsung Magician doesn’t seem to recognize them. Thanks for your help.

    • blank

      It is my understanding that the EVOs were never meant to be used in RAID configurations, which is probaby why the MAgician software is not recognizing them. For that reason, and better reliability, I chose to stick with the 840 Pro and 850 PROs.

      • blank

        owner of a 840 Pro, if you raid them you can’t see them in magician and trim is always disabled(stripe 0,1 etc) <<—no matter what model you have, same applies to 850 PRO

      • blank

        Trim has been supported for ssd raid arrays since 7 series motherboards (z77).. download the intel raid driver for your os and your golden. (driver can only be installed once in raid mode)

      • blank

        NOT in raid 0,1,5 ..ONLY raid mode but as single disks (TRIM is useless if you use RAID 0(stripe), or raid 1(mirror))

        Do your homework and come back 😉

        P.S. IDE, AHCI, RAID modes but nowdays only AHCI and RAID modes available

      • blank

        article clearly states that trim is indeed supported for raid 0 arrays
        on 7 series motherboards. Ive already done my “homework” sir.

      • blank

        Well that’s great, in the past the only thing intel did was to manage single drives in “raid mode” if they manage raid0 in 2 drives then kuddos to intel. Too bad you need specific motherboards to do that..

    • blank

      For me, I would set up which ever way is easiest for you. I would personally set them up in RAID 0 and store both the OS and programs on the same volume. Just make sure you do routine backups and you should be golden. Samsung Magician does not detect the drives as you have noticed, it is normal. I personally don’t bother with installing Magician in my own system.

    • blank

      my friend has a gaming pc with 2 1tb ssds in raid0 he said its the fastest storage he has seen (and he is a professional it technician). but it all depends on the type of raid controller. if its virtual raid I don’t think you should do it unless you really want all your space to be unified. but if its hardware raid you would probably get a faster write and read times.

      just my opinion

    • blank

      Unless you are in an enterprise scenario. The raid 0 doesn’t affect real life performance, as most the time the SSD’s bandwidth isn’t your bottleneck. I have done a raid 0 with 850 pros and looked into upgrading my 850 pro to a 950 pro. Benchmark wise yeah the speed performance is way higher. But when i comes to boot time and application speed there is very little difference. Plus with raid 0 your data isn’t safe, and with Rapid Mode’s low latency, it becomes more of a want then a should. shabammm

  3. blank

    Hi, so if the RAPID tech is ON, then I can get some performance gain in PC games for example or not really? 🙂 (I mean if these games are installed to this SSD & also this games are very often loading textures & other datas from this SSD where the RAPID tech is ON.)

    • blank

      Definitely not if you’re using a gaming computer. Textures are loaded into VRAM (memory on the video card itself) which is much faster than system RAM or the SSD.

      Even machines without a video card will cache as much texture data as possible in System RAM.

  4. blank

    I have 28 GB RAM on a HP Envy 700 430 qe. When I use the rapid drive, I get scores of 3700+ MB Read and 2400+ MB Write. Is that real? The computer is very fast with the 850 Pro 512.

    • blank

      In RAPID mode you are basically using a RAM drive that gets synchronized to disk as quickly as possible. Your RAM is very much faster than the SSD. But you risk losing the data in RAM if your computer loses power.

      • blank

        That’s what a laptop battery is for. To have no interruption when mains power is lost. I have a business class HP Workstation with two batteries on it for double security. On the workstation there is a lot going on to insure data loss is avoided. It is the perfect environment for the RAPID Mode, especially with 32GB RAM installed and a second SSD in the Optical Drive Bay. The way I work on a laptop makes data loss minimal too. Any data I lose will likely be progress since the last save. I am contantly saving my progress and so it will not be much. I also copy and paste so there is always a the original on the drive if the copy is interrupted – unlike with cut and paste when all the info is in RAM. It is tempting to cut a 29GB (Windows has 3GB reserved) chunk of info and instantly paste it and I don’t. RAPID mode pretty much does this and in a safer way. With all my interfaces SATA III running at 6MB/Sec and with 2 USB3.0 ports and a Sata port, transfering data is nifty, especially if I pack the info into a single zip instead of transferring individual files.

        Backup transfers from the main SSD C: drive to the secondary SSD in the optical drive bay is is not noticed.

        Where I see a big difference is any batch file processing, any repetitive process and during Photoshop CS6 Suite tasks which use only two Cores instead of all 4 because of sloppy programming and design, RAPID mode improves the speed. I have four 50gb scratch disk partitions on my 2TB C: Drive to double the processing speed of CS6 and any for any other program which needs to cache lots of temporay info on a separate partition.

        The claim it can help boot up times is rubbish. Rapid Mode loads after windows has started up.


    • blank

      I have only 4GB RAM (DDR2) on a 6+ years old Lenovo Laptop (Core2Duo). The SSD is connected via SATA2 port. Even I got 1800 MB/s read & 1300 MB/s write speed, when rapid mode is enabled. But practically there was no other benefit other than the benchmark results (Other benchmarking tool revealed similar result too). So, finally, I disabled it as it took too much RAM space from my 4GB RAM.

      • blank

        You don’t really ‘lose’ that RAM, it’s actually a small RAM disk speeding along your computer. Even one with a SATA-1 MB would benefit, that is, it’s it’s an AHCI enabled one. Samsung tends to be picky, Intel SSD’s less so.

        The RAPID app only uses 20-25% of your available RAM, depending on if it’s the 840 or 850 series, and in real world testing, I’ve found the RAPID app help SATA-2 computers more so than SATA-3, which usually has other built in features to boost performance.

        With 4GB, at the most, RAPID will use 1GB RAM, and take my word for it, the RAM is NOT wasted. Your browsers, open tabs, everything you do is being boosted by RAPID. Plus your SSD gets less wear & tear, which doesn’t mean a lot these days as long as it’s aligned & not defragged, making it work less & your lower cost RAM do it’s job.


  5. blank

    RAPID mode causes crashes–Windows 7,SP1, Samsung 850EVO. I enabled RAPID mode two months ago and ever since I have seen weird hangs and crashes every three or four days. Disabled RAPID two weeks ago and have seen no crashes since.

  6. blank

    So with Rapid mode for the average user gain is…none, the same time you lose more than 1GB of RAM which is far more valuable for Gaming or Multitasking. Nice feature but I wont ever use it.

    • blank

      Good job I have 16gb of ram. Wooohooo I will use it all the time!!!!

    • blank

      32GB of memory here – this is really good for me.

    • blank

      you don´t lose anything.. if the memory is needed it will be released by the rapid software.
      rapid only uses FREE resources.
      if your game needs all the system memory it can get…then the rapid cache will be reduced to zero.
      but it´s a useless feature anyway.
      only good for PR.

      • blank

        I feel that’s a bit of an untrue statement, it does not release it. running games on very high settings with a gtx 1060 6gb of ram paired with 16gb of system and the games simply crash when trying to open up other programs while alt tabbed giving an out of memory error, upon inspection of win10 task manager it shows around 7 gigs inside the cache at the time of crash when it said it was out.

        So if that was the case it would of released whatever was inside the 7 gigs of cache and allowed more memory to run whatever i was alt tabbed trying to do.

        Thus what led me here in search of do I really want samsung to take up around 25% of my ram for—-? what gain in performance? and where. Far as I can see it’s a compromise if you don’t have TONS of ram, (32-64gigs), In this case trading multitasking for speed.

  7. blank

    don’t care if the performance gain is minimal, I have 32gb ram 😛

  8. blank

    Wouldn’t those with SATA-2 based computers see as much as, if not more benefit, than one on a fast SATA-3 computer? Intel’s Rapid Storage Technology App is already built into many of these fast SATA-3 PC’s, so that’s why I see SATA-2 based computers as potentially benefitting the most from RAPID. Modern computers with fast controllers should need this as much as slower ones, which needs all of the help available.

    Just a thought. I use the RAPID feature on my three SATA-2 based computers and see & feel a difference, whereas on my more powerful SATA-3 based PC with a 500GB Samsung 850 Pro, I really can’t tell the difference. In fact, seems to make the PC act funky more than anything, so disabled it & it’s naturally fast again. It’s not that way on my SATA-2 computers, the speed difference is there.


  9. blank

    Rapid mode uses a different algorithm than Windows cache. It caches on block level, while Windows caches on file level. File level usually has more information, that’s why in most scenarios it’s better than block level hashing. But there are a few use cases where block level caching could be beneficial. For example, you have a virtual machine. From Windows point of view, it’s one monolithic 30GB file and Windows won’t cache it. But in fact, data moves within that file, and there is some pattern to these movements. A good block caching software can pick up that pattern and improve performance. Some other specific scenarios involving big data files can benefit too. But most Windows users won’t see any benefit.

  10. blank

    still works perfectly on windows 10. just needs a little workaround to get it running

  11. blank

    the fazit is pretty much nonsense.
    you may can benchmark latency differences but the performance gain of RAPID (only under some circumstances) are so small that you won´t notice them EVER.
    i would call myself a poweruser.
    i am using premiere, nuke, 3d max, photoshop, after effects, solidworks.
    i am always looking to make my system faster.. but rapid is not doing anything for me.
    my own testing showed no benefits of rapid mode at all.
    beside making low level benchmarks look better.

  12. blank

    But what does this mean for games? there was BF3 and WoW listed but nothing was tested, is it more beneficial inside the games load times to have it enabled?, i know this is an old post just was a good read so had to ask.

  13. blank

    Rapid can cause latencies of up to 31ms, which is a potential disaster for real-time usage, such as audio recording.

    Most applications benefit from having MAXIMAL MB per second transfers speeds, but real time applications rely upon MINIMAL kB per millisecond deliveries to be responsive.

    Maximum throughput is only achieved by running buffers full most of the time, which means EVERTHING is delayed.

    It’s basic queuing theory. You get served more quickly at a bank if there is no queue, but the bank wants better teller utilisation which it gets by ensuring that there are always queues.

    The drive manufacturers are like the banks in wanting greatest throughput for impressive specs, whereas your usage may require the fastest response. You need to make sure you choose the right drive criteria for your usage.

Leave a Reply

Your email address will not be published. Required fields are marked *