To add to this, all of your RAM will only work as fast as the slowest RAM. So in the above case where you have 500Mhz RAM and 4000Mhz RAM, all sticks would be limited to 500Mhz. This is something to always keep in mind when upgrading. Buy matching RAM whenever possible. If you're getting all of it at once, try to get RAM made in matching batches.
@piddlefoot cute.... Lets compare Apples to apples here, and not mix it up with arbitrary numbers. Obviously there are limits to hardware. My point is: currently I am running @ 1333mhz, the low end on ram speeds. The difference from that to 2133 is negligible. Sure there's a diff...... but who is really going to notice the diff? Run your benchmarks, and tell me how much more FPS you get. Please do share your findings, as it would be news if it was greatly significant. I mean, if you like spending money for bragging rights, go ahead. If you feel your rig is giving you an advantage, all the power to you.
Faster RAM = Faster load times and smoother play. That's all there is to it. I tend to trust personal experience over some faceless youtuber any day of the week, plus you can find youtube videos to support any ignorant theory like flat Earth, Moon landings being faked, etc.
This absolutely 100% needs to be fixed. Performance degrades over time. SC examples: After flying around the planet once (102fps): Spoiler After logging in same spot (121fps): Spoiler Settings were maxed, 1080p. This is 100% repeatable every single time I play the game, and if I fly around the planet a 2nd time, I'd lose another 15-20fps, etc. Until my frame rates are sub 60. This is an absolutely crippling performance/optimization issue. It keeps me from really playing the game. Also, in the same area where the SCs were taken, there was an area below the base, where I could take 10 steps, and see a 55 frame rate swing. Relogged into the game, and that particular frame rate variation between spots was gone. Some really, really odd performance hitches in this game. Lots of incredibly massive frame rate swings over short distances. If we're not getting beta any time this year, then maybe you can at least address some of the worst optimization issues before then? Thanks. Empyrion could, and should run consistent high frame rates on my PC. 120fps+ PC specs: CPU: Ryzen 5 1600x @ 3.9ghz Ram: 16GBs DDR4 3200mhz GPU: GTX 1080ti SSDs: 480GB Sata III + 256GB PCI-E Windows 10 64 bit. Also, I get I/O related load stuttering when flying around, even on my Samsung EVO 960 256GB SSD which is around 3x faster than my Sata III SSD is. That's just insane. Why is this happening? Show me one more game in existence where a Samsung EVO 960 gets I/O stuttering? I ran latencymon in the background while playing empyrion, just in case there was any possibility that it was something else causing the stuttering, some driver related issue, etc, and everything was deep in the green.
Ram speeds have a negligible impact on Empyrion. I've been testing ram speeds as well as timings on my PC. Difference between 2400mhz and 3200mhz is about 2-3fps in Empyrion. And that's with the same CAS timings! If I tightened the timings up on 2400mhz, the difference would be closer to 1-2fps. I'm thinking about backing my ram down to 2933mhz, because 3200mhz on this kit requires 1.42v, and 2933mhz runs stable at 1.37v. There has been a lot of talk about ram speeds on these new Ryzen CPUs, and based on my testing, I'm not seeing the impact in performance a lot of these people on the internet are claiming. Then of course, I don't know what the motives are of these people putting out benchmarks and graphs on the internet, like if they're getting paid to cook some numbers up and make it look presentable, just to get suckers to overpay for ram, etc. All I can say is that their testing doesn't match up with mine. And I'm not telling people they should go out and buy premium ram kits. When ram prices are already through the roof.
As was said, it is not about a higher frame rate, it's about smoother play and faster loading time as you move around making it more enjoyable. Higher frame rates do not equal more enjoyable. I have never used any AMD hardware so I don't really care about that aspect as it does not affect me. I also don't care about benchmarks, I never play benchmarks for entertainment value as they have none. I only care about actual gameplay as it pertains to my game experience personally. YMMV.
@crimsonedge885 Sounds like you have done your homework, and went the distance to explore the reasons. I agree, there is a thing such as overkill & over paying for hardware that doesn't provide that edge over others. Some like to ignore fact because it conflicts with their beliefs; That's cognitive dissonance. Ultimately, smooth and stable fps surely has an affect on the experience/enjoyment factor.
Cognitive dissonance is having inconsistent thoughts or beliefs as it relates to your behaviors and choices, for example if you believe more RAM will do nothing for a computer and but the maximum anyways. It is not staying with a single consistent belief and practicing actions that are in line with that belief. That is called being true to yourself. As for me, the size of the files I generate require huge amounts of RAM just to be able to manipulate and edit those files, so I have the most RAM that will fit a given system.
Cognitive dissonance occurs when your ideas, beliefs, or behaviors contradict each other: if, for example, you see yourself as smart but can't believe you made such dumb stock investments. How we choose to resolve the dissonance (and its accompanying discomfort) is a reflection of our mental health.
The test on RAM is simple, put some old 400mhtz RAM against a stick of 3000mhtz RAM and tell me theres no difference. Its a very simple scientific formula that the PC industry have improved on for decades now. People may argue all they like about 1000mhtz vs 2000mhtz makes bugger all difference in there games, makes NO difference to the science behind RAM speeds. People fail to take into account any games code will screw any tests you try, because no two games are the same, the ONLY way to test RAM is on a test specifically designed for RAM, period. Testing RAM speeds via what a game does is flawed and will NEVER give you any kind of scientifically accurate data about the actual RAMs ability. Thats just testing what the ability of that game is, and it may not handle the operations in the most optimal way, its code may favour one RAM set of timings over another and other anomalies pop up also, testing RAM via a game is a very very rough guesstimate scientifically speaking. Thats like saying copper cables are just as good for the internet, as Optical Cable is.........Speed is everything in gaming. Hows that for a crappy analogy. Just because you cant tell the difference in a game between RAM speeds from equal capacity amounts of RAM, doesnt mean there isnt an actual gain, its just another example of your failure to do a proper scientific measurement. Better explained. The speed rating of your RAM module is an expression of its data transfer rate. The faster the number, the faster your computer can store and retrieve the data stored in local memory. The formula for the exact speed rating changes slightly based on the version of DDR memory your computer is using . It’s no longer simply an expression of clock speed, like a processor, but a combination of hardware factors. But in general, faster is better. Though the speed rating is usually expressed in straight “DDR” terms, we also have the old PC2/PC3/PC4 standard still hanging on. These numbers generally follow the speed rating corresponding to the generational standard: “DDR3 1600 RAM” is also labelled as “PC3 12800,” “DDR4 2400 RAM” is also “PC4 19200,” and so on. This is a technicality based on the old bit and byte data expression—one byte equals eight bits. So, if the first number is DDR 1600, expressed in million bytes per second capability, the second number is PC3 12800, expressed in million bits per second. 12800 divided by eight is 1600, so it’s two ways of stating the same thing. Generally, things will be less confusing if you stick to the first “DDR2/3/4” speed rating. In addition to standard speed ratings, each RAM module also has a rating for something called timings. This is expressed as a series of four numbers, like 5-5-5-15 or 8-8-8-24. We’re getting into some advanced computer science topics here, dealing with the specific amount of time it takes the module to access single bits of data across columns and rows of the memory array. But for the sake of brevity, this collection of numbers is generally referred to as “latency.” Latency deals with how fast the RAM module can access its own hardware, and in this specific case, the lower the numbers, the better. Lower latency means faster data access, thus faster data transfer to the CPU, and faster operation of your computer overall. Higher-quality, more expensive RAM has lower latency, and both this rating and the RAM’s clock speed can be overclocked by enthusiasts. That being said, the differences in latency are so minuscule that unless you’re running industry-level server operations or multiple virtual machines, you’re unlikely to see any real difference between RAM with a higher or lower latency.
Not sure what versions these all are (Running A8 EXP atm), but so far I've been getting frame drops down to 8-15fps every day/night cycle change for a good 3-5 min until things normalize again. (36-49fps avg.) And I'm running mid to low graphics settings. As for 7.6.1 Stable...I can run close to high on everything and still get 80+ frames using an i3 Dual core and a GeForce GTX 750Ti. While the game as a whole is still in Alpha stage, the stable version is WAY more optimized. I like the changes as much as anyone, but when it drops 15-20 frames out of the blue, just because of a day cycle change... I'm going to stick with playing scenarios on 7.6 until they finally get A8 stable, and then see if it's worth switching back to.
It's a known bug and they're working on it... from the A8 Exp3 changelog Changes: - Increased Scenario upload size limit from 10 to 100 MB - New setting for pole biomes in playfield.yaml > see how "Biome: Pole" is now defined for eg TemperateStarter - Adapted Planet Axis in dynamic.yaml for all playfields > now between -20 and 20 - Improved: FPS drop at dusk & dawn in Oasis on desert planet > pls send us save games if you find areas with heavy FPS drops
I OC'd the ram on my PC, and noticed no difference in frame rates between 2400mhz ram, and 3200mhz ram. There are games like Fallout 4 where there is a difference, but Empyrion isn't one of those games. 90%+ of games, ram speed has a negligible impact on frame rates and performance. I don't think anyone has benchmarked this game and posted their results. Intel vs AMD CPUs, DDR3 vs DDR4 ram, Nvidia vs AMD GPUs, SSD vs HDD, 8GBs of ram vs 16GBs vs 36GBs of ram, etc. Maybe the developers could weigh in on this? Assuming they use multiple test rigs of a variety of specifications? Results could be spread like this: - Highest frame rate - Average frame rate - Bottom 1% low frame rate (lots of things could potentially have an impact on this, drive speeds, ram amount, etc) - Frame timings - Frame rate stability
You are misinterpreting my words, I never said RAM will give you higher framerates. I said it affords smoother gameplay, ie, less lagging and jitters as areas and resources load. I* think you may have glossed over a lot of the conversation judging by your comment.
FWIW I get ~35 fps running at 4k with a 1080ti, 6700k (stock speed) and 16GB memory, with OS and steam on an SSD, with graphics on high settings. Not amazing, but fine for an early release game. It's not like it's unplayable. However, I agree that when I downscale to 1080, it doesn't rock and roll quite as fast as one might predict indicating a need for further optimization, but I'm not concerned since that optimization will likely come once alpha is done and we're into beta. Too much other stuff changing right now.
Another one of these "game blah blah runs 1000x faster than Empyrion so there must be something wrong with Empyrion" posts. It's like the complaints for another similiar game called 7D2D. You do realize Emp is a completely destroyable / voxel based game compared to 99% of the other games out there and therefore you cannot compare performances between them. Its much easier to calculate and render a Fallout4 for example, compared to Empyrion. I'm sure there will be optimizations when the game reaches gold status but I also wouldn't expect it to be massively better than what it currently is due to the requirements of the current engine.
I've heard the "b....bu.....but!!!! It's voxels!" argument as a defense for poor optimization way too many times, and at this point I just don't care about the excuses anymore. Meaning it has zero basis on my opinion about the matter. All I know is that I cannot recommend this game to a single person, until it's optimized according to reasonable standards. I recommended to a few friends to not buy this game until it hits beta, because they'll probably just end up refunding it. 8.0 experimental runs a little better for me. So it has improved a little bit, but I do expect considerable optimization improvements to be made in the future. At least get rid of the load hitching on SSDs, if they can't manage to fix that, then I'll be extremely disappointed. No game should ever require that much speed to choke out a SSD. I've heard a few people comment that this game has to be on a ram disk to not hit an I/O bottleneck, and that's insanely un-optimized. Voxels or not, that's unacceptable. If I had known this game used Unity Engine prior to purchasing, I wouldn't have bought it. I think the optimization issues have more to do with Unity Engine being complete garbage for optimization than it being a voxel game. Lesson learned. Don't buy any more Unity Engine games. Every Unity Engine game I've played has really bizarre, annoying performance related issues. Load hitching is pretty common in Unity Engine games, even when running on SSDs.
