• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Should i buy primocache?

Status
Not open for further replies.
Joined
Jun 22, 2021
Messages
68 (0.06/day)
System Name Samsung Galaxy Book3 Pro 14
Processor i7 1360P
Motherboard -
Cooling Dual Fan Cooling System
Memory 16 GB 6000MHZ DDR5
Video Card(s) Intel Iris Xe G7 96EUs
Storage 512 GB NVMe SSD
Display(s) 1800p display 120HZ
Case -
Audio Device(s) Quad Speakers (2x 5W Woofers, 2x 2W Tweeters)
Power Supply 65 watt
Mouse Hp omen gaming mouse
Keyboard -
VR HMD -
Software Windows 11
Benchmark Scores Geekbench 6 : single core: 2252 multicore : 10456 compute(OpenCL):16734 compute(Vulcan):21841
it tripled my SSD performance!
 

Toothless

Tech, Games, and TPU!
Supporter
Joined
Mar 26, 2014
Messages
9,449 (2.50/day)
Location
Washington, USA
System Name Veral
Processor 5950x
Motherboard Asus Crosshair VIII Hero Wi-Fi
Cooling Corsair H150i RGB Elite
Memory 2x16GB G.Skill TridentZ
Video Card(s) Powercolor 7900XTX Red Devil
Storage Crucial P5 Plus 1TB, Samsung 980 1TB, Teamgroup MP34 4TB
Display(s) Acer Nitro XZ342CK Pbmiiphx + 2x AOC 2425W
Case Fractal Design Meshify Lite 2
Audio Device(s) Blue Yeti + SteelSeries Arctis 5 / Samsung HW-T550
Power Supply Corsair HX850
Mouse Corsair Nightsword
Keyboard Corsair K55
VR HMD HP Reverb G2
Software Windows 11 Professional
Benchmark Scores PEBCAK
It didn't. It's using ram as a cache. Your drive isn't faster..
 

Mussels

Freshwater Moderator
Staff member
Joined
Oct 6, 2004
Messages
58,413 (8.08/day)
Location
Oystralia
System Name Rainbow Sparkles (Power efficient, <350W gaming load)
Processor Ryzen R7 5800x3D (Undervolted, 4.45GHz all core)
Motherboard Asus x570-F (BIOS Modded)
Cooling Alphacool Apex UV - Alphacool Eisblock XPX Aurora + EK Quantum ARGB 3090 w/ active backplate
Memory 2x32GB DDR4 3600 Corsair Vengeance RGB @3866 C18-22-22-22-42 TRFC704 (1.4V Hynix MJR - SoC 1.15V)
Video Card(s) Galax RTX 3090 SG 24GB: Underclocked to 1700Mhz 0.750v (375W down to 250W))
Storage 2TB WD SN850 NVME + 1TB Sasmsung 970 Pro NVME + 1TB Intel 6000P NVME USB 3.2
Display(s) Phillips 32 32M1N5800A (4k144), LG 32" (4K60) | Gigabyte G32QC (2k165) | Phillips 328m6fjrmb (2K144)
Case Fractal Design R6
Audio Device(s) Logitech G560 | Corsair Void pro RGB |Blue Yeti mic
Power Supply Fractal Ion+ 2 860W (Platinum) (This thing is God-tier. Silent and TINY)
Mouse Logitech G Pro wireless + Steelseries Prisma XL
Keyboard Razer Huntsman TE ( Sexy white keycaps)
VR HMD Oculus Rift S + Quest 2
Software Windows 11 pro x64 (Yes, it's genuinely a good OS) OpenRGB - ditch the branded bloatware!
Benchmark Scores Nyooom.
Aww, i was hoping this would be a good primocache thread.

I have a frakton of spare RAM and played with it, but want to see other peoples thoughts and advice on how best to use it before i pay.
 
Joined
Jul 5, 2013
Messages
26,119 (6.47/day)
Aww, i was hoping this would be a good primocache thread.

I have a frakton of spare RAM and played with it, but want to see other peoples thoughts and advice on how best to use it before i pay.
Ok, I'll be fair, if you have 32GB+ and you leave at least 16GB for your system, this is useful, but 32GB is the bare minimum! There-in lays the catch, you have to have a serious amount of RAM to get the most out of this program and it is very much a case of the-more-memory-the-merrier!

The program is found below;
Learn more here;
And the Specs;
The minimum is stated as 2GB, and the recommend 4GB, but there is no way in hell a user will make the most of what this program can do unless you give it huge chunks of RAM. The useful minimum is 32GB system RAM with 16GB allocated to PrimoCache.

Honestly, the prices are very reasonable(lifetime use) if you can make the most of it.
 
Last edited:
Joined
Jun 22, 2021
Messages
68 (0.06/day)
System Name Samsung Galaxy Book3 Pro 14
Processor i7 1360P
Motherboard -
Cooling Dual Fan Cooling System
Memory 16 GB 6000MHZ DDR5
Video Card(s) Intel Iris Xe G7 96EUs
Storage 512 GB NVMe SSD
Display(s) 1800p display 120HZ
Case -
Audio Device(s) Quad Speakers (2x 5W Woofers, 2x 2W Tweeters)
Power Supply 65 watt
Mouse Hp omen gaming mouse
Keyboard -
VR HMD -
Software Windows 11
Benchmark Scores Geekbench 6 : single core: 2252 multicore : 10456 compute(OpenCL):16734 compute(Vulcan):21841
thanks, i have 20gb ram
 

Mussels

Freshwater Moderator
Staff member
Joined
Oct 6, 2004
Messages
58,413 (8.08/day)
Location
Oystralia
System Name Rainbow Sparkles (Power efficient, <350W gaming load)
Processor Ryzen R7 5800x3D (Undervolted, 4.45GHz all core)
Motherboard Asus x570-F (BIOS Modded)
Cooling Alphacool Apex UV - Alphacool Eisblock XPX Aurora + EK Quantum ARGB 3090 w/ active backplate
Memory 2x32GB DDR4 3600 Corsair Vengeance RGB @3866 C18-22-22-22-42 TRFC704 (1.4V Hynix MJR - SoC 1.15V)
Video Card(s) Galax RTX 3090 SG 24GB: Underclocked to 1700Mhz 0.750v (375W down to 250W))
Storage 2TB WD SN850 NVME + 1TB Sasmsung 970 Pro NVME + 1TB Intel 6000P NVME USB 3.2
Display(s) Phillips 32 32M1N5800A (4k144), LG 32" (4K60) | Gigabyte G32QC (2k165) | Phillips 328m6fjrmb (2K144)
Case Fractal Design R6
Audio Device(s) Logitech G560 | Corsair Void pro RGB |Blue Yeti mic
Power Supply Fractal Ion+ 2 860W (Platinum) (This thing is God-tier. Silent and TINY)
Mouse Logitech G Pro wireless + Steelseries Prisma XL
Keyboard Razer Huntsman TE ( Sexy white keycaps)
VR HMD Oculus Rift S + Quest 2
Software Windows 11 pro x64 (Yes, it's genuinely a good OS) OpenRGB - ditch the branded bloatware!
Benchmark Scores Nyooom.
thanks, i have 20gb ram
Let's just assume 16GB is needed for the OS and programs, as thats a common recommendation

That means you've got 4GB of data cached, and you cant choose what that data is

Are you really going to benefit? Are you reading/writing small data so often and so fast, that your SSD cant keep up?
 
Joined
Jan 31, 2012
Messages
2,558 (0.56/day)
Location
East Europe
System Name PLAHI
Processor I5-10400
Motherboard MSI MPG Z490 GAMING PLUS
Cooling 120 AIO
Memory 32GB Corsair LPX 2400 Mhz DDR4 CL14
Video Card(s) PNY QUADRO RTX A2000
Storage Intel 670P 512GB
Display(s) Philips 288E2A 28" 4K + 22" LG 1080p
Case SilverStone Raven RV03
Audio Device(s) Creative Sound Blaster Z OEM
Power Supply Fractal Design IntegraM 650W
Mouse Logitech Triathlon
Keyboard REDRAGON MITRA
Software Windows 11 Home x 64
Man, I had forgotten about this one. There was a time I had played with it, but it was a while back. It all depends of course. If you have decent SSD/ RAM perhaps AV software (if needed), backup software (paid if needed) and you still feel the need to invest into your PC, yeah go for it.

EDIT: what I mean is, I have way more other bases to cover before I even consider splashing cash on this kind of software.
 
Joined
Jul 5, 2013
Messages
26,119 (6.47/day)
thanks, i have 20gb ram
Ok, try this since you already have the trial version: Set your cache allocation for 8GB. That will leave 12GB for your system, which is a doable amount of RAM. You can always try 4GB as well if your system performance dips.
 
Joined
Jul 16, 2014
Messages
8,157 (2.23/day)
Location
SE Michigan
System Name Dumbass
Processor AMD Ryzen 7800X3D
Motherboard ASUS TUF gaming B650
Cooling Artic Liquid Freezer 2 - 420mm
Memory G.Skill Sniper 32gb DDR5 6000
Video Card(s) GreenTeam 4070 ti super 16gb
Storage Samsung EVO 500gb & 1Tb, 2tb HDD, 500gb WD Black
Display(s) 1x Nixeus NX_EDG27, 2x Dell S2440L (16:9)
Case Phanteks Enthoo Primo w/8 140mm SP Fans
Audio Device(s) onboard (realtek?) - SPKRS:Logitech Z623 200w 2.1
Power Supply Corsair HX1000i
Mouse Steeseries Esports Wireless
Keyboard Corsair K100
Software windows 10 H
Benchmark Scores https://i.imgur.com/aoz3vWY.jpg?2
I have a frakton of spare RAM
In a wooden frame hanging on the wall? :D :rolleyes::nutkick:

On topic ~
While I was play testing a mod for 7 Days to Die, ram usage topped 27Gb, took a while to load. Those with less than 16gb systems experience really crap performance. Using a tool like this would put that whole thing into the pagefile, could ruin a drive lol.
 
Last edited:
Joined
Sep 17, 2014
Messages
21,588 (6.00/day)
Location
The Washing Machine
Processor i7 8700k 4.6Ghz @ 1.24V
Motherboard AsRock Fatal1ty K6 Z370
Cooling beQuiet! Dark Rock Pro 3
Memory 16GB Corsair Vengeance LPX 3200/C16
Video Card(s) ASRock RX7900XT Phantom Gaming
Storage Samsung 850 EVO 1TB + Samsung 830 256GB + Crucial BX100 250GB + Toshiba 1TB HDD
Display(s) Gigabyte G34QWC (3440x1440)
Case Fractal Design Define R5
Audio Device(s) Harman Kardon AVR137 + 2.1
Power Supply EVGA Supernova G2 750W
Mouse XTRFY M42
Keyboard Lenovo Thinkpad Trackpoint II
Software W10 x64
Never had any trust in these solutions. It just reeks too much of downloading RAM or 'your computer has 670161 problems we can fix'.

As if the most efficient data management isn't already built into the whole rig as it is...
 
Joined
Mar 10, 2010
Messages
11,878 (2.26/day)
Location
Manchester uk
System Name RyzenGtEvo/ Asus strix scar II
Processor Amd R5 5900X/ Intel 8750H
Motherboard Crosshair hero8 impact/Asus
Cooling 360EK extreme rad+ 360$EK slim all push, cpu ek suprim Gpu full cover all EK
Memory Corsair Vengeance Rgb pro 3600cas14 16Gb in four sticks./16Gb/16GB
Video Card(s) Powercolour RX7900XT Reference/Rtx 2060
Storage Silicon power 2TB nvme/8Tb external/1Tb samsung Evo nvme 2Tb sata ssd/1Tb nvme
Display(s) Samsung UAE28"850R 4k freesync.dell shiter
Case Lianli 011 dynamic/strix scar2
Audio Device(s) Xfi creative 7.1 on board ,Yamaha dts av setup, corsair void pro headset
Power Supply corsair 1200Hxi/Asus stock
Mouse Roccat Kova/ Logitech G wireless
Keyboard Roccat Aimo 120
VR HMD Oculus rift
Software Win 10 Pro
Benchmark Scores 8726 vega 3dmark timespy/ laptop Timespy 6506
Aww, i was hoping this would be a good primocache thread.

I have a frakton of spare RAM and played with it, but want to see other peoples thoughts and advice on how best to use it before i pay.
I payed , I don't use it now, simply not worth the effort for a typical home user.
Games didn't all of a sudden fly or load much quicker.
 
Joined
Mar 15, 2021
Messages
63 (0.05/day)
depends on the use case - just for read cache for SSD probably no, it's good for cache/prefetch on HDD (ofc Windows does that too but with Primocache you have control over it) and great if you need functions like write L2 caching and defer write function
 
Joined
Jul 5, 2013
Messages
26,119 (6.47/day)
Decided to actually test this rather than rely on past experiences. Testing showed interesting results. See screenshots.
The testing system is one of my Dell T3500s with 12GB of RAM and a Xeon W3680 6core CPU running Windows 11 Build 22000.194. I wanted to use this system as it's the one I've been testing Windows 11 on and has the least amount of RAM of all my desktop type systems.

This is a test on a non-cached partition, which shows the bare drive performance. This acts as a control baseline result. This is what was expected from the drive in question(Seagate 7200RPM HDD).
NonCachedDrive.jpg


This is with a 2GB cache. As you can see, there is a marked improvement in read speeds, but not much in write speeds, which were only slightly better.
2GBCache.jpg


This is with a 3GB cache. This is where the read speeds bump an even larger margin(more than double). The write speeds went back down to the baseline, not sure why.
3GBCache.jpg


This is a 4GB cache. Once again an improvement, but not a big one compared to the 3GB test.
4GBCache.jpg


This is with a 6GB cache. This was an interesting improvement. The memory bandwidth of that system is 13GB per second so the cache is approaching the system bandwidth limit.
6GBCache.jpg


So I was wrong, this program is not a waste. This software is not like others of this type. However, the testing clearly shows that you need a fair amount of system RAM to use it effectively.

@OctupleGolf001
You should be able to allocate 6GB of your system RAM and not choke your system while taking advantage of what this program has to offer. I'm changing my advice to: Yes, buy it if you will use it alot. Seems to work really well.

@Everyone
I think the pattern in the results is clear here, the more RAM the better, but unlike disk caching programs of the past, a frak-ton of RAM in NOT actually needed. You can get a solid result just using a 2GB cache.

My original recommendation though still stands, the more RAM the better. The advice I'm changing is the bare minimum, which is 12GB. I will say if you're going to use this program on a constant basis, start with 24GB and an 8GB cache. RAM is affordable these days so upgrading if you need to should not be expensive. This program seems to be very efficient and 8GB cache size should provide anyone a great result.
 
Last edited:

silentbogo

Moderator
Staff member
Joined
Nov 20, 2013
Messages
5,491 (1.41/day)
Location
Kyiv, Ukraine
System Name WS#1337
Processor Ryzen 7 3800X
Motherboard ASUS X570-PLUS TUF Gaming
Cooling Xigmatek Scylla 240mm AIO
Memory 4x8GB Samsung DDR4 ECC UDIMM
Video Card(s) Inno3D RTX 3070 Ti iChill
Storage ADATA Legend 2TB + ADATA SX8200 Pro 1TB
Display(s) Samsung U24E590D (4K/UHD)
Case ghetto CM Cosmos RC-1000
Audio Device(s) ALC1220
Power Supply SeaSonic SSR-550FX (80+ GOLD)
Mouse Logitech G603
Keyboard Modecom Volcano Blade (Kailh choc LP)
VR HMD Google dreamview headset(aka fancy cardboard)
Software Windows 11, Ubuntu 20.04 LTS
Primocache is essentially superfetch on steroids. Short-term satisfaction is good (so as "on paper" numbers), but actual daily benefits are highly questionable, especially if you are already running an SSD.
I assume Windows 11 only feels snappier only because they fixed preloading algos to account for higher RAM capacities, which means it's another con to buying software which can potentially become obsolete in a couple of months.
 
Joined
Dec 5, 2013
Messages
628 (0.16/day)
Location
UK
Aww, i was hoping this would be a good primocache thread.

I have a frakton of spare RAM and played with it, but want to see other peoples thoughts and advice on how best to use it before i pay.
I tried a 16-24GB RAMDisk on a 32GB RAM system, and quite honestly it isn't worth it for general gaming. My findings:-

1. If you switch your computer off every night then you're basically going to have to "install" / copy the game over anyway every day. Total loading time = copy time from HDD/SSD to RAMdisk + the time when you start the game from the RAMDisk + the extra "human time" you spend setting it up is actually longer than just running it off an SSD.

2. Exponentially depreciating gains. An 80% reduction from say 75s (5,400rpm HDD) to 15s (SSD) feels huge. But a further 5% reduction from 15s (SSD) to 14.3s (RAMDisk) is not noticeable in the real world, especially when 11s of both involves unskippable intro movies, epilepsy warnings, etc, or CPU bottlenecked decompressing data files. I have one smallish 1.6GB game (Quantum Conundrum) that takes 34s to load on a 5,400rpm HDD and yet still 29-30s on a RAMDisk / SSD. Why? It's an endless string of intro movies you can't skip. RAMDisks won't help with this stuff.

3. Whenever you access any file, you're loading it into the Windows file cache anyway. So when you copy a file from SSD to RAMDisk, you're actually making two copies : 1. RAMDisk and 2. Windows file cache. And the first place Windows looks is the latter. So for many games you're not actually starting it from the RAMDisk but the cache. If you have a lot of RAM (16-32GB) and your game only uses say 6GB, then Windows will be holding it all in RAM anyway even without a RAMDisk even if you made a "null" file copy or simply "read" the files some other way, eg, MD5-ing the game folder will cache a game into RAM (Windows File Cache) even without a RAMDisk.

4. Games are getting larger than RAMDisks can handle faster than RAM is coming down in price. With +50-100GB game installs becoming more common, it's already getting pretty ridiculous and expensive to buy 64-128GB RAM to try and keep up with the "rat race" vs buying an increasingly cheap 2TB SSD.

Bottom line : If you never reboot your system, play only the same small game over & over, day after day, can't afford a large SSD yet simultaneously happen to have +32GB of expensive RAM lying around unused there's probably some value in RAMDisks to some people. For most people, however, there isn't. Overall, I decided a 2TB SSD was a better investment than RAMDisk-ing 32GB RAM and a whole lot less fiddling about "pre-loading" every time I rebooted / shut down / wanted to play something else. The only real practical use I found for a RAMDisk was a 1GB web browser cache for heavy daily browsing sessions, and even then most of that benefit was less about speed and more about reducing SSD defragmentation / maximising PE cycles.

Edit: I know Primocache works differently than a pure RAMDisk (most commonly / recently accessed based algorithm) but the same principle applies - if you play only the same small games over & over and perhaps have a fast boot SSD + a large secondary HDD, it may increase the speed of games loaded from the latter. But if you play a variety of games (especially larger ones), then whatever the algorithm decided to cache / pre-load yesterday may not be what you want to play today and it's no different to loading from an SSD directly.
 
Last edited:
Joined
Jul 5, 2013
Messages
26,119 (6.47/day)
Primocache is essentially superfetch on steroids.
No, this seems to be something else entirely. The caching is on RAM not on disk like SuperFetch is.
Short-term satisfaction is good (so as "on paper" numbers), but actual daily benefits are highly questionable
I have not done any long term tests, but I predict that there is something tangible to gain.
especially if you are already running an SSD.
This I doubt, but it can be tested just as easily on the same system which has an SSD installed. I'd be willing to run tests with that drive if requested.

I assume Windows 11 only feels snappier only because they fixed preloading algos to account for higher RAM capacities, which means it's another con to buying software which can potentially become obsolete in a couple of months.
Those are a few assumptions. Time will tell.
I tried a 16-24GB RAMDisk on a 32GB RAM system, and quite honestly it isn't worth it for general gaming. My findings:-
That is because your configuration was flawed. If you have 32GB of RAM and you set a 16GB or 24GB cache, your system performance will tank as a result. The rest of your points are invalid as they are based on your experience with a flawed configuration. Try an 8GB or 12GB cache and let us know how it goes.
 

silentbogo

Moderator
Staff member
Joined
Nov 20, 2013
Messages
5,491 (1.41/day)
Location
Kyiv, Ukraine
System Name WS#1337
Processor Ryzen 7 3800X
Motherboard ASUS X570-PLUS TUF Gaming
Cooling Xigmatek Scylla 240mm AIO
Memory 4x8GB Samsung DDR4 ECC UDIMM
Video Card(s) Inno3D RTX 3070 Ti iChill
Storage ADATA Legend 2TB + ADATA SX8200 Pro 1TB
Display(s) Samsung U24E590D (4K/UHD)
Case ghetto CM Cosmos RC-1000
Audio Device(s) ALC1220
Power Supply SeaSonic SSR-550FX (80+ GOLD)
Mouse Logitech G603
Keyboard Modecom Volcano Blade (Kailh choc LP)
VR HMD Google dreamview headset(aka fancy cardboard)
Software Windows 11, Ubuntu 20.04 LTS
No, this seems to be something else entirely. The caching is on RAM not on disk like SuperFetch is.
The only difference is that Superfetch does it automagically without user intervention, while PrimoCache creates a RAM disk(and wastes resources) to do essentially the same thing manually.
Most linux distros do similar thing with preload. The only thing that's kinda beneficial - is user control(e.g. choosing what to preload), other than that - I can't really think of any.
 
Joined
Dec 5, 2013
Messages
628 (0.16/day)
Location
UK
That is because your configuration was flawed. If you have 32GB of RAM and you set a 16GB or 24GB cache, your system performance will tank as a result. The rest of your points are invalid as they are based on your experience with a flawed configuration. Try an 8GB or 12GB cache and let us know how it goes.
I tried all sizes (starting with 2GB) and ended up at 16-24GB (for caching sub 10GB smaller games outright). It can make a huge difference if you're loading from a 5,400rpm HDD and the algorithm gets it right, but vs a fast SSD, it didn't made much difference in the real-world for reasons mentioned. My RAMDisk shows CrystalDiskMark Sequential screenshots of over 12,000MB/s vs only 520MB/s from an MX500 SATA SSD, yet that really doesn't mean games will load 23x faster as games simply don't load like that. In the real world, they'll request a chunk of data, then go through a lot of CPU cycles unpacking it, initialising it, etc, only then request another chunk, etc. It isn't a continuous "sequential" stream at all. "If my SSD is 2,000MB/s and my RAMDisk is 10,000MB/s then that means game load times will fall from 15s to 3s because that's what CrystalDiskMark says" as some believe just doesn't happen in the real world for the same reason the newest P31 NVMe review isn't showing 6x lower load times for games than 2016-era SATA's on the same chart. The theoretical CrystalDiskMark Sequential -> real-world Game Load times never was something that ever scaled linearly, let alone do so to infinity beyond depreciating gains of which fast SSD's are already well into the realms of.
 
Last edited:
Joined
Feb 7, 2021
Messages
12 (0.01/day)
System Name Evangeline
Processor Ryzen 7 5700X
Motherboard ASRock B550 Taichi Razer Edition
Cooling bequiet Shadow Rock LP / Arctic P12 PWM PST RGB 0Db
Memory 2x32GB DDR4-3200 18-20-20 Patriot Viper Blackout
Video Card(s) Nvidia RTX A4000
Storage Crucial P5+ 1TB + MEGA Fastro MS200 2TB + Crucial MX500 1TB
Display(s) Acer VG280K + Dell S2721QS
Case SilverStone Grandia GD09
Audio Device(s) Asus Xonar U7 mkII, Audio Technica ATH-GDL3
Power Supply BeQuiet Straight Power 11 1000W
Mouse Roccat Kone AIMO
Keyboard Keychron Q6 / Durock Lupine switches / Generic keycaps
Modern SSDs already use a technology called HMB (Host Memory Buffer) that uses system RAM for performance optimization, even aside from any preloading and caching your OS does for you, so paying money for yet another thing that puts RAM between your Operating System and your SSD seems of marginal use at best.
 
Joined
Jul 5, 2013
Messages
26,119 (6.47/day)
The only difference is that Superfetch does it automagically without user intervention, while PrimoCache creates a RAM disk(and wastes resources) to do essentially the same thing manually.
Most linux distros do similar thing with preload. The only thing that's kinda beneficial - is user control(e.g. choosing what to preload), other than that - I can't really think of any.
I tried all sizes (starting with 2GB) and ended up at 16-24GB (for caching sub 10GB smaller games outright). It can make a huge difference if you're loading from a 5,400rpm HDD and the algorithm gets it right, but vs a fast SSD, it didn't made much difference in the real-world for reasons mentioned. My RAMDisk shows CrystalDiskMark Sequential screenshots of over 12,000MB/s vs only 520MB/s from an MX500 SATA SSD, yet that really doesn't mean games will load 23x faster as games simply don't load like that. In the real world, they'll request a chunk of data, then go through a lot of CPU cycles unpacking it, initialising it, etc, only then request another chunk, etc. It isn't a continuous "sequential" stream at all. "If my SSD is 2,000MB/s and my RAMDisk is 10,000MB/s then that means game load times will fall from 15s to 3s because that's what CrystalDiskMark says" as some believe just doesn't happen in the real world for the same reason the newest P31 NVMe review isn't showing 6x lower load times for games than 2016-era SATA's on the same chart. The theoretical CrystalDiskMark Sequential -> real-world Game Load times never was something that ever scaled linearly, let alone do so to infinity beyond depreciating gains of which fast SSD's are already well into the realms of.
I'm not going to argue. Believe what you want.

Moving on...
So after the mention of SSDs, I decided to test the difference. Once again the results were interesting... Same system, just on the Windows 7 installation on an SSD. Here we go..

NonCached partition. Results were as expected.
NonCached.jpg


2GB cache. Once again, the read speeds jumped up dramatically but had a different mix of results from the HDD.
2GB.jpg


3GB cache. Simply massive improvement! This is effectively my system RAM bandwidth minus processing over-head.
3GB.jpg


4GB cache. Almost no change from 3GB. Margin of error flucutation kind of thing.
4GB.jpg


6GB cache. Again, little change VS 3GB.
6GB.jpg


What can we conclude from this series of results? For one, this program offers a much better performance spread with an SSD VS HDD. The second point is that the performance ceiling is limited by overall system RAM bandwidth, so the faster your system RAM, the better the performance. Third, with an SSD, the need for a large cache seem unnecessary, whereas with an HDD a larger cache is of benefit.

I am now very curious how this software actually functions. It seems to have it's own unique caching scheme.

Thinking it's time for a test on a more powerful system with larger data sizes..

EDIT;
More tests done. Tried this on a different system, one with a Xeon E5-2667v2 & 32GB of RAM and an SSD. Once again the results were interesting.

Non-Cached partition.
NonCached.jpg


2GB cache. The boost wasn't the same as the other system's SSD 2GB run. Weird.
2GB.jpg


3GB cache.
3GB.jpg


4GB cache.
4GB.jpg


6GB cache.
6GB.jpg


8GB cache.
8GB.jpg


I did 12GB and 16GB runs with little difference.

I'm going to run some benchmarks later just to compare load times. I know of at least one game engine based benchmark that accounts for drive load times. Will post later...
 
Last edited:
Joined
Feb 20, 2020
Messages
9,340 (5.77/day)
Location
Louisiana
System Name Ghetto Rigs z490|x99|Acer 17 Nitro 7840hs/ 5600c40-2x16/ 4060/ 1tb acer stock m.2/ 4tb sn850x
Processor 10900k w/Optimus Foundation | 5930k w/Black Noctua D15
Motherboard z490 Maximus XII Apex | x99 Sabertooth
Cooling oCool D5 res-combo/280 GTX/ Optimus Foundation/ gpu water block | Blk D15
Memory Trident-Z Royal 4000c16 2x16gb | Trident-Z 3200c14 4x8gb
Video Card(s) Titan Xp-water | evga 980ti gaming-w/ air
Storage 970evo+500gb & sn850x 4tb | 860 pro 256gb | Acer m.2 1tb/ sn850x 4tb| Many2.5" sata's ssd 3.5hdd's
Display(s) 1-AOC G2460PG 24"G-Sync 144Hz/ 2nd 1-ASUS VG248QE 24"/ 3rd LG 43" series
Case D450 | Cherry Entertainment center on Test bench
Audio Device(s) Built in Realtek x2 with 2-Insignia 2.0 sound bars & 1-LG sound bar
Power Supply EVGA 1000P2 with APC AX1500 | 850P2 with CyberPower-GX1325U
Mouse Redragon 901 Perdition x3
Keyboard G710+x3
Software Win-7 pro x3 and win-10 & 11pro x3
Benchmark Scores Are in the benchmark section
thanks, i have 20gb ram
Hi,
If you fill in your system spec's everyone that post on your threads... would already know this and more

Asus ROG boards come with ramcache included
I've personally never used it some buddies use it mainly for benchmarks and some use it for programs
Vulnerabilities execute through memory so it's a double edged sword feature like virtual machine/ virtualization which I always disable one less security hole.
 
Joined
Dec 5, 2013
Messages
628 (0.16/day)
Location
UK
I'm not going to argue. Believe what you want.
There's nothing to "believe". I literally just tested it (1. Picked a game at random, Mark of the Ninja (and it's the GOG offline installer version so no client running spending time doing DRM checks), 2. Start it from a 5,400rpm HDD then a SATA SSD then a RAMDisk (with a cold boot in-between each to clear the Windows Cache) and 3. Record "time from start click until reaching the main menu":-

A. (5,400rpm HDD) = 12.1s
B. (MX500 SSD) = 3.9s
C. (RamDisk) = 3.8s

^ And that's with the entire game pre-cached (a 100% perfect cache prediction hit rate of a game that completely fits into RAM) vs a slow (by modern standards) SATA SSD. For larger games that won't fit into a RAMDisk, or perhaps playing something else the algorithm hasn't cached, the 3.9 vs 3.8s would be reduced even more as it'll be loading direct from the SSD anyway. As I said, regardless of the Quattuordecillion Yottabytes per FemtoSecond CrystalDiskMark sequential marketing screenshots, real world game load times = depreciating gains...

What can we conclude from this series of results?
That unless you "play" CrystalDiskMark all day long, you're not actually testing anything of real-world substance beyond how CrystalDiskMark specficially can saturate multi GB/s loads in a way normal games / applications don't. (Hint: If you're looking at this software for the purpose of reducing game load times, have you tried testing some actual games)?... ;)
 
Joined
Jul 5, 2013
Messages
26,119 (6.47/day)
If you fill in your system spec's everyone that post on your threads... would already know this and more
https://www.techpowerup.com/forums/account/specs
This is NOT a requirement and not everyone looks.

There's nothing to "believe". I literally just tested it (1. Picked a game at random, Mark of the Ninja (and it's the GOG offline installer version so no client running spending time doing DRM checks), 2. Start it from a 5,400rpm HDD then a SATA SSD then a RAMDisk (with a cold boot in-between each to clear the Windows Cache) and 3. Record "time from start click until reaching the main menu":-

A. (5,400rpm HDD) = 12.1s
B. (MX500 SSD) = 3.9s
C. (RamDisk) = 3.8s

^ And that's with the entire game pre-cached (a 100% perfect cache prediction hit rate of a game that completely fits into RAM) vs a slow (by modern standards) SATA SSD. For larger games that won't fit into a RAMDisk, or perhaps playing something else the algorithm hasn't cached, the 3.9 vs 3.8s would be reduced even more as it'll be loading direct from the SSD anyway. As I said, regardless of the Quattuordecillion Yottabytes per FemtoSecond CrystalDiskMark sequential marketing screenshots, real world game load times = depreciating gains...


That unless you "play" CrystalDiskMark all day long, you're not actually testing anything of real-world substance beyond how CrystalDiskMark specficially can saturate multi GB/s loads in a way normal games / applications don't. (Hint: If you're looking at this software for the purpose of reducing game load times, have you tried testing some actual games)?... ;)
You are missing important context and as I said...
I'm not going to argue.
...so move along.
 
Joined
Jan 5, 2006
Messages
18,428 (2.72/day)
System Name AlderLake / Laptop
Processor Intel i7 12700K P-Cores @ 5Ghz / Intel i3 7100U
Motherboard Gigabyte Z690 Aorus Master / HP 83A3 (U3E1)
Cooling Noctua NH-U12A 2 fans + Thermal Grizzly Kryonaut Extreme + 5 case fans / Fan
Memory 32GB DDR5 Corsair Dominator Platinum RGB 6000MT/s CL36 / 8GB DDR4 HyperX CL13
Video Card(s) MSI RTX 2070 Super Gaming X Trio / Intel HD620
Storage Samsung 980 Pro 1TB + 970 Evo 500GB + 850 Pro 512GB + 860 Evo 1TB x2 / Samsung 256GB M.2 SSD
Display(s) 23.8" Dell S2417DG 165Hz G-Sync 1440p / 14" 1080p IPS Glossy
Case Be quiet! Silent Base 600 - Window / HP Pavilion
Audio Device(s) Panasonic SA-PMX94 / Realtek onboard + B&O speaker system / Harman Kardon Go + Play / Logitech G533
Power Supply Seasonic Focus Plus Gold 750W / Powerbrick
Mouse Logitech MX Anywhere 2 Laser wireless / Logitech M330 wireless
Keyboard RAPOO E9270P Black 5GHz wireless / HP backlit
Software Windows 11 / Windows 10
Benchmark Scores Cinebench R23 (Single Core) 1936 @ stock Cinebench R23 (Multi Core) 23006 @ stock
Should i buy primocache?

Sure if you got the cash for the cache....
 
Status
Not open for further replies.
Top