This is especially true for people who live in places with high ambient temperatures.
Is it worth to upgrade CPU?
- Thread starter alteration
- Start date
-
We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
This is especially true for people who live in places with high ambient temperatures.
You can find great value builds online, copy it down and go to a local computer shop, put in an order for all the parts and pay an assembly fee. Last time I did that here in oz (4 years and 2 computers ago), it was a $70 AUD (approx $50 USD) assembly fee. Plus I got it covered by warranty because all the parts and the assembly came from the one shop.
Overclockers do some nice bundles if you don't want the hassle of set up.I think its time to upgrade from my old i5 2005 which I have had for ages.
Remember to check Hardwareswap on Reddit! You can find amazing deals on there too.
You are partially right, but in a round-about way. In Stellaris there a ton of computations that can be done in parallel. Ships moving inside each and every starsystem only need to account for the local ships in the same system. Every empire sector has an independent economy and AI associated with it. Every combat only account for the fleets actively involved in it. These things can all run in parallel, and a lot of them don't need to run on every single game day. But in order for a gameday to be "done", you actually need everything for that day to have completed computation.
The slowest tasks have in the past been huge sectors, where the AI would spend a lot of time evaluating all possible actions across all systems under it's control. Second on the list is huge fleet combats. Both of these can be multithreaded, but are simpler to develop and test as single-threaded logic. There are most likely also a lot of potential optimizations even within the logic, but they all require developertime and apparently, there are more pressing issues to work on
Same problem here.
i5 6600
GTX 970
16gb Ram
Around 1000 pops fps drops from around 180 at the start of the game to 30 now. Really takes the fun away. Playing on a medium map with 2 Fallen empires and 3 other ai empires.
No problem in any other pdx game and here i wanted to conquer the whole galaxy :/. I hope Utopia fixes some of the fps problems till then i will shelf the game
i5 6600
GTX 970
16gb Ram
Around 1000 pops fps drops from around 180 at the start of the game to 30 now. Really takes the fun away. Playing on a medium map with 2 Fallen empires and 3 other ai empires.
No problem in any other pdx game and here i wanted to conquer the whole galaxy :/. I hope Utopia fixes some of the fps problems till then i will shelf the game
One of the devs recently confirmed that the AI is already running massively paralell. However, there is limits to that too: At some point the main game tick has to wait for aIl AI's to make thier decisions before the next game tick. Otherwise we end up with an AI doing nothing while the game runs fluid.
It is that delay of the next game tick that we humans will notice.
I have a good computer and I'm also having lag problems with this game. Not only with Stellaris. For example with Shogun 2 which I still play there are optimization issues where the GPU goes mad. Stellaris is not hungry about system requeriments, graphics are very simple compared to other games. It's more about optimization and making efficient use of computer resources than upgrading your hardware. So in the end I think that with a better computer you won't notice a massive change in this game performance.
My specs are: AMD-FX, 8Gb ram, Radeon R9 and a corsair SSD. Early game it goes smooth but mid-late game everything changes. Just give time to PDX, sooner or later they are going to find what is causing this problem.
My specs are: AMD-FX, 8Gb ram, Radeon R9 and a corsair SSD. Early game it goes smooth but mid-late game everything changes. Just give time to PDX, sooner or later they are going to find what is causing this problem.
The AMD-FX 8150 isn't a very good CPU for gaming. The Bulldozer CPUs suffered from poor single thread performance, and not many games use 8 threads well enough to compensate for that. The R9-200 is a series consisting of cards like the R9-270, R9-280, and the R9-290. I recall that the R9-290 was about on par with the GTX 970, but I don't think the 270 or 280 are nearly as good. You could probably find out what exact card you have by using GPU-Z.
Unfortunately I don't think those specifications qualify for even a midrange gaming PC these days.
Sorry man, I just figured it would be better knowing what you have hardware-wise so you know what's bad optimization or simply your PC.
However, have you thought about overclocking your FX 8150? I've heard those CPUs can hit around 4-4.3 GHz if you have a good cooling solution, which could net a nice performance increase. This is assuming that your using a PC and not a laptop, and that you know how to do overclocking properly.
Last edited:
CPUs aren't a bottleneck unless he's playing on some silly tiny resolution like 640x480. Stellaris isn't an especially CPU or GPU intensive game anyway. Someone mentioned their i5 getting 40-45 C, which is similar to what my AM2+ Phenom II Thuban hits playing Stellaris.
And no, an overclocked Groundshaker won't be much better than an overclocked Pentium IV TBH. Both P4 and Earthmover were both "Speed Demon" type uarches, focusing on "Phat Megahurtz" and Big Airflow to do stuff. Raisin/Grape is AMD's first "Brainiac" design, comparable to Intel Core (the first Intel Brainiac design was Nehalem) from the early 2010s.
AMD continues to overfocus on multithreading. Great for workstations (these days?) but mediocre for games. Raisin promises great multithreading performance and decent single core performance, though, which is fantastic for people who use PCs for real work that isn't just "vidya gams", but also for video/photo/3D editing and other stuff that takes advantage of Big Threads.
Also it's ICE ICE BABY compared to the world's first commercial thermonuclear reactor Bulldozer/Excavator/Piledriver/Steamroller. Unfortunately RX 480 retains the thermonuclear capability of the Earthmovers. We'll have to wait for Vega to see ICE ICE BABY hit the GPU scene. Maybe by 2025 AMD will reach the apex of silicon development that Intel has hit and release a good single-core Brainiac design that has cold temps and combines AMD's experience with multithreading and learning how to do single-cores. Just in time for Intel or IBM to release the first commercial photonic uarch and make all silicon obsolete.
tl;dr CPUs aren't a serious gamer thing.
And no, an overclocked Groundshaker won't be much better than an overclocked Pentium IV TBH. Both P4 and Earthmover were both "Speed Demon" type uarches, focusing on "Phat Megahurtz" and Big Airflow to do stuff. Raisin/Grape is AMD's first "Brainiac" design, comparable to Intel Core (the first Intel Brainiac design was Nehalem) from the early 2010s.
AMD continues to overfocus on multithreading. Great for workstations (these days?) but mediocre for games. Raisin promises great multithreading performance and decent single core performance, though, which is fantastic for people who use PCs for real work that isn't just "vidya gams", but also for video/photo/3D editing and other stuff that takes advantage of Big Threads.
Also it's ICE ICE BABY compared to the world's first commercial thermonuclear reactor Bulldozer/Excavator/Piledriver/Steamroller. Unfortunately RX 480 retains the thermonuclear capability of the Earthmovers. We'll have to wait for Vega to see ICE ICE BABY hit the GPU scene. Maybe by 2025 AMD will reach the apex of silicon development that Intel has hit and release a good single-core Brainiac design that has cold temps and combines AMD's experience with multithreading and learning how to do single-cores. Just in time for Intel or IBM to release the first commercial photonic uarch and make all silicon obsolete.
tl;dr CPUs aren't a serious gamer thing.
Last edited:
GPU can not be the bottleneck. If anything CPU delays interfere with drawing passes and thus FPS more.
It only uses 4 GiB memory, so RAM can not be it either.
CPU bottleneck without heating up is realy easy to explain:
If it taxes out one core while leaving the others more or less idle, the total heat of the CPU will not increase. The colling is designed for all cores being taxed 50-100%.
That was both the reason they started to and the reason they could so easily add build-in overclocking: Cooling designed for 4 cores at 3 GhZ full load can easily cool one core at 5 GhZ.
Someone doesn't play on a high enough resolution, then.
CPUs rarely increase performance in games substantially. You will always be better off upgrading to a new GPU, unless your mainboard doesn't support it, in which case you'll need a new CPU, unless by happenstance your new mainboard supports your old socket I guess. They're basically tertiary upgrades after mainboard and GPU if you want to increase your frames per second.
GPUs are almost always the bottleneck in frames. Games simply aren't big enough number crunchers to tax CPUs very much, unless you have a strange case like you're crunching PhysX on your CPU instead of your GPU (for whatever silly reason). Stellaris isn't a very intense game anyway. An Ivy Bridge will crush it into submission as much as a Skylake in any realistic measurement, so a new CPU is not going to be a panacea for any sort of end game FPS drop. That's inherent to the version of Clausewitz engine that Stellaris runs on.
Last edited:
It is worth a upgrade if your CPU is mediocre like the AMD-FX 8150. If you have a good CPU with good single thread performance it's almost always better off to upgrade to a new GPU rather than a better CPU. But if you do not, your CPU will bottleneck your graphics card especially if your running something like a GTX 1070 or GTX 1080. Simply turning up the resolution won't matter if your CPU doesn't have enough performance keep up with your graphics card, which will lead to a decrease in FPS.
For instance putting a GTX 1080 with something like a Pentium 4 is ludicrous, and a CPU like that will hold your graphics card back. The CPU is the processing power of the PC and if it can't keep up with more powerful hardware then the whole system will fall behind.
When you get to CPUs like a modern I5 or I7 you'll rarely get a GPU bottleneck, but older processors running on older architecture with far less single core performance? Your definitely going to want to upgrade your CPU if your GPU is any good.
Or if the GPU is really old, Intel's onboard GPU might even be more powerful which would be a nice bonus.
And you have it backwards. Stellaris is a very CPU intensive game. Just because your processor doesn't max at 100% across the board doesn't mean it isn't CPU intensive. When the CPU shows high usage (using 100% as an example) it could mean your CPU is actually overloaded. Stellaris is actually far less GPU intensive than CPU intensive, like every other Paradox game using the Clauswitz Engine. In fact, the current lag problems are because of the fact the game is a bit too CPU intensive (lack of optimization), which is why lag will occur later game due to having so much data to process.
Things like this code loop brings even the best CPU to it's knees.
Last edited:
I have the same CPU, i5 3570k, but I have it overclocked to 4.5 GHz. Try bumping it up to that and see if there's a difference.
I have a Lenovo Y50-70 laptop (i7-4710HQ 2.5 GHz, Nvidia GTX 860m, 16GB ram, 512 GB SSD). My longest run was 200 years -- on 1.3 -- and I haven't experienced any lag. Although I was playing on fast and normal speeds most of the time, I don't know whether game speed has anything to do with the day's end calculation.
Your statement is true for 3D games like DOOM, For Honor etc but not even close to correct for games on the Clausewitz engine. Telling someone to upgrade his GPU for this game is about the worst advice possible.
If my experiences a patch or two ago are anything to go by, an Ivy will not crush this game, an Ivy will result in lag late game, just like all available CPU's do. NO CPU you can buy is sufficiently fast to avoid lag. As others have said, just because the game doesn't make the CPU report 100 percent in task monitor, doesn't mean it's not the bottleneck. If you want a game that can be crushed by any decent CPU, then you need to look at something like Tyranny or Overwatch.
I don't know what this post has to do with the OP but a couple of points
-What the hell does the energy efficiency of an RX480 have to do with someone who has gtx 1080?
-the reason AMD has a design that is better for servers than gamers is simple, they designed a server chip and also used it to sell to gamers. The only reason Naples is not out yet is because it takes longer to verify server chips than consumer ones
-single core performance is not very far behind skylake/kaby lake, the main issue is the clock deficit not the ICP. This is most likely due to the low power process they manufacture it on-something that will pay dividends in the mobile space. That they are this close is kind of amazing for a company that is so far behind in research dollars but probably irrelevant to the thread.
So you just repeated what I said, but veiled it as a disagreement?
Ivy crushes this game as much as a Skylake does. The bottleneck is literally the engine. A Skylake might see minor improvements but not substantial ones. You're much better off getting a Z170/Z270 chipset for the modern interface ports than for any hoped for improvements in late game FPS drop.
FX 8150 will probably be about as good as my Thuban is, which is sufficient until you get about 150-200 years into the game, and then it becomes slowed down to about normal speed. I'm not expecting much better performance with a Ryzen 1700 either, because the bottleneck is literally the Clausewitz engine. Until that's improved, you're not going to see much performance gains in buying a new CPU.
Upgrading to a modern chipset is more valuable than upgrading the CPU itself, sans thermal management of Raisin or something, because modern interface ports are important for peripherals like GPUs and HDDs/SSDs. DDR4 compatibility is also nice. These benefits alone are enough to invest in a new chipset of either Raisin or Sky/Kabylake, but don't expect FPS to improve very much.
1) It's not better "for servers", it's better for workstations. Computers that do real work, like 3D modeling, video editing, and streaming of data. Stuff that has been optimized for multithreading for a long time because its more valuable in productive/wealth creation aspects of computing than playing video games.
2) Single thread performance is quite far behind Skylake/Kaby Lake in games, as much as 20-40% behind. It's essentially Ivy Bridge at that point. This is hugely impressive compared to Earthmover-series and more than adequate for any modern video game, unless you are playing on a very low resolution like 720p. It's quite adequate for 1080p.
3) Yeah it's literally a once in a century style comeback. Too bad investors hyperfocus on game performance (like video games matter LoL) and completely ignored the massive ass blasting Raisin gave Skylake in real things like content creation and workstation uses.
If AMD plays this right, it might start pushing Intel from the much more valuable/larger workstation market. Unfortunately I doubt GlobalFoundries has the production capacity to do that.
Last edited: