hdds RAM Cache placebo?

[SpodermanBR]

Ok guys, so I recently came across Primo RAM Cache which sounded a bit whacky to me (it was advertised to me by a random youtube video) I got the trial version and tested it out. Turned out it did reduce load times in some games (Destiny 2) when I had it running with around 512mb allocated. So I went and did some tests on CrystalDiskMark "real world performance" profile. It seems a bit off to me...
Check this out, this is it with RAM cache (640Mb on disk C:, and some other smaller amounts on the other partitions, its a 2 HDD setup with 2 parts on HDD1 and 1 part on HDD2)

 

Now this is it without the RAM cache running

Testing a file transfer from HDD2 to HDD1 without the RAM cache gave me around 55MB/s (windows explorer MB) peaked at 75MB/s for some seconds and dipped at 28 twice, 12gb file.

 

Now with the RAM cache active in both HDDs...

Less speed but more consistent at 50MB/s, while having peaks at 66MB/s and dips around 48 minimum.

Can someone tell me why Destiny 2 load faster with the cache on and the transfer rate is still more or less the same?

Without cache when I double click the icon it takes 30 seconds to open, 1 minute to get to the logo screen (where you press enter) and 4 minutes to get to my character screen.
With cache its 30 seconds to open, 44 seconds to get to the logo screen and 1 minute and 20 seconds to get to my character screen.
Tested it three times and the results vary by 1-2 seconds more or less.

[Electronics Wizardy]

So this software is doing some bad things that makes benchmakrs looks good, but can cause issues with data loss.

 

So windows already has a ram cache, its just there is a mode to disable this that crystal disk mark and most other benchmarks use. That way you can see the actual speed of the drive, not of your ram.

 

THe other issue is that it lets files say they have been written when there really not, if you lose power, you can get a corruped disk as the filesystem expects that some files are on disk when there really aren't

 

 

[SpodermanBR]

4 minutes ago, Electronics Wizardy said:

So this software is doing some bad things that makes benchmakrs looks good, but can cause issues with data loss.

 

So windows already has a ram cache, its just there is a mode to disable this that crystal disk mark and most other benchmarks use. That way you can see the actual speed of the drive, not of your ram.

 

THe other issue is that it lets files say they have been written when there really not, if you lose power, you can get a corruped disk as the filesystem expects that some files are on disk when there really aren't

 

 

I enabled defer-write to test the higher write speeds, but nothing different and I'm running on a laptop so power loss isn't a big issue. But the question remains, why does Destiny 2 have a difference in load time?

[SkilledRebuilds]

But you already explained it  .

Destiny 2 uses dara from a faster pool of data and your wondering why its faster.

 

Latency and File Access times would be MUCH lower.

 

Even if a HDD and  SSD were limited to 50MB/s the SSDZ WINS due to how quickly it can find that data. (Latency per file)

[Mihle]

That HDD you have is slow for a HDD.

[Escapenz]

When i done it was great for me, The again i gave it 40gb of ram L3 cache and 1tb ssd L2 cache 

[SpodermanBR]

1 hour ago, Mihle said:

That HDD you have is slow for a HDD.

Oh yea, it's an older laptop HDD (I'm thinking 2013), still faster than most laptop HDDs I've had though. I'm not accustomed to high storage speeds other than my UFS 2.1 smartphone, my PC still runs 2 HDDs at 7200RPM. Where I live (Brazil) the cost of basically anything is rather expensive and electronics is much more expensive than usual (abusive taxes and whatnot), so when I built my system I focused more on the essentials, which were still used parts and cost a bit over 2x the minimum wage. I'm currently looking to sell my CPU/RAM/MOBO kit for cheap and get a newer Ryzen one, but I'm struggling find good deals in quarantine.

 

 

8 hours ago, SkilledRebuilds said:

But you already explained it  .

Destiny 2 uses dara from a faster pool of data and your wondering why its faster.

 

Latency and File Access times would be MUCH lower.

 

Even if a HDD and  SSD were limited to 50MB/s the SSDZ WINS due to how quickly it can find that data. (Latency per file)

Ok, so I'll keep this running with a little bit of RAM allocated to the disk corresponding to the game. My laptop has 8Gb 1600MHz DDR3 and I usually don't use a lot, plus, I still manage to get little over 50fps on Destiny with settings on low, that's fine by me I just want to have this as an option when I'm on university and have some free time.

 

1 hour ago, Escapenz said:

When i done it was great for me, The again i gave it 40gb of ram L3 cache and 1tb ssd L2 cache 

Impressive, I've only got 12Gb on my PC, as I mentioned in the beginning of this reply, I'm trying to upgrade my kit and currently thinking of getting 16Gb RAM on my next one. I think it's enough for most games and while I may have to do some analyses and statistics for future science stuff, I've done it before with 8Gb and it worked out fine.

Anyway, thanks for all the replies, I guess this bags up the question.

[RejZoR]

RAM cache is pointless. Windows already does that to some extent. SSD cache however does A LOT. Mostly for read caching, but can also be used for write caching tho be prepared for much faster wear of the SSD cache when used for write cache too. Then again small SSD's go for peanuts these days and it'll still last several years before dying of wear.