|
It doesn't even have that much going on for it, graphically. I don't think anyone has really figured out what's wrong with it, it's just horrendously optimized.
|
#
?
Oct 25, 2023 01:55
|
|
|
|
| # ? Jun 8, 2024 20:30 |
|
|
They probably spent a lot of time thinking about CPU efficiency and much less about GPU efficiency, because yeah sim game, and then when they realized how deep the hole was they didn’t have time to dig out before shipping. (Because so many things in game publishing are predicated on “when you release” than “what you release” for reasons that are probably historically interesting but totally beyond my ken.)
|
|
#
?
Oct 25, 2023 02:12
|
|
|
Happy to see the 2400G still has legs. Love that guy - got me through 2020
|
|
#
?
Oct 25, 2023 02:21
|
|
|
repiv posted:i don't know why we need two or more relatively slow CPU threads when the GPU thread lumps everything together and works fine, but IIRC there was pushback to a CPU megathread last time it came up tbf a CPU thread is also a sort of a motherboard and RAM thread and it sort of makes sense to not mix up intel's X420 with AMD's X420. but also, Subjunctive posted:It’s OK. We have 4 Ukraine threads, we can have this war in multiple threads too.
|
|
#
?
Oct 25, 2023 02:23
|
|
|
CPUs should have multiple threads. It's called hyperthreading 
|
|
#
?
Oct 25, 2023 02:32
|
|
|
gradenko_2000 posted:CPUs should have multiple threads. It's called hyperthreading boo
|
|
#
?
Oct 25, 2023 03:22
|
|
|
gradenko_2000 posted:CPUs should have multiple threads. It's called hyperthreading 
|
|
#
?
Oct 25, 2023 04:14
|
|
|
um actually this is the amd thread, so it’s SMT  *piano falls directly on my head*
|
|
#
?
Oct 25, 2023 04:26
|
|
|
gradenko_2000 posted:CPUs should have multiple threads. It's called hyperthreading by that logic we should have thousands of GPU threads 
|
|
#
?
Oct 25, 2023 04:33
|
|
|
Klyith posted:by that logic we should have thousands of GPU threads Don't you mean CUDA™ posts™?
|
|
#
?
Oct 25, 2023 05:17
|
|
|
BlankSystemDaemon posted:Why is a sim game GPU bottlenecked? In one twitch clip I saw, turning off depth of field increased the fps from like 20 to 110.
|
|
#
?
Oct 25, 2023 05:24
|
|
|
gradenko_2000 posted:CPUs should have multiple threads. It's called hyperthreading 
|
|
#
?
Oct 25, 2023 09:53
|
|
|
BlankSystemDaemon posted:Why is a sim game GPU bottlenecked? What if they're trying to offload some of that simulation onto the GPU?
|
|
#
?
Oct 25, 2023 10:11
|
|
|
Everything I've seen of Skylines 2 that isn't the pre-rendered trailer makes it look the same graphically as the first, so IDK what they've been doing.
|
|
#
?
Oct 25, 2023 11:43
|
|
|
apparently some of the post processing passes are horrendously slow, just disabling depth of field is doubling peoples frames
|
|
#
?
Oct 25, 2023 11:47
|
|
|
gradenko_2000 posted:CPUs should have multiple threads. It's called hyperthreading Cygni posted:um actually this is the amd thread, so it’s SMT Kibner posted:In one twitch clip I saw, turning off depth of field increased the fps from like 20 to 110. isndl posted:What if they're trying to offload some of that simulation onto the GPU? BlankSystemDaemon fucked around with this message at 12:43 on Oct 25, 2023 |
|
#
?
Oct 25, 2023 12:32
|
|
|
BlankSystemDaemon posted:Why is a sim game GPU bottlenecked? Cities skylines performed terrible on launch, was never really fixed, and benefitted from hardware improvements. Cities skylines 2 seems to have followed in its footsteps.
|
|
#
?
Oct 25, 2023 14:10
|
|
|
not only are the post processing passes ridiculously expensive, it also sounds like they're just drawing too much geometry https://store.steampowered.com/news/app/949230/view/3744239011016556921?l=english
|
|
#
?
Oct 25, 2023 17:02
|
|
|
VostokProgram posted:not only are the post processing passes ridiculously expensive, it also sounds like they're just drawing too much geometry quote:It is worth mentioning that for a game like this, the performance target is to run at a steady 30FPS minimum. There is no real benefit in a city builder to aim for higher FPS (unlike a multiplayer shooter) as a growing city will inevitably become CPU-bound. What matters more with this type of game is to avoid stutters and have a responsive UI. when I first booted the game, it was literally running at 10fps in the main menu on my 7800X3D and 4090. Just rendering a simple skyline in the background. Great job nailing that target, guys.
|
|
#
?
Oct 25, 2023 17:23
|
|
|
what raised my eyebrows in that article is that they can't implement TAA, which means they can't implement FSR / DLSS
|
|
#
?
Oct 25, 2023 17:24
|
|
|
what raised my eyebrow in their first statement and hasn't been improved in that clarification are their implications about simulation and ui logic not being entirely decoupled with priority given to the ui... it's impressive to look at this release compared to cities skylines 1's release against sim city loving everything up, i presume they've had a complete change in employees since then
|
|
#
?
Oct 25, 2023 17:30
|
|
|
What's funny is that there is a TAA option in the game, but I'm not sure if it actually does anything. I managed to get 60fps at a native 4K just by turning off depth of field, and setting volumetrics and GI to medium. But that's with an early city, and it really shouldn't take a 4090 to do that given how unremarkable the game looks. It's an absolute shimmery mess with SMAA.
|
|
#
?
Oct 25, 2023 18:03
|
|
|
Dr. Video Games 0031 posted:What's funny is that there is a TAA option in the game, but I'm not sure if it actually does anything. I managed to get 60fps at a native 4K just by turning off depth of field, and setting volumetrics and GI to medium. But that's with an early city, and it really shouldn't take a 4090 to do that given how unremarkable the game looks. It's an absolute shimmery mess with SMAA. I dont know if the TAA actually works, but switching from SMAA to TAA in the advanced settings makes the game look massively better with no discernible change in framerate. Game is 6-12 months underbaked.
|
|
#
?
Oct 25, 2023 18:06
|
|
|
gradenko_2000 posted:what raised my eyebrows in that article is that they can't implement TAA, which means they can't implement FSR / DLSS
|
|
#
?
Oct 25, 2023 19:22
|
|
|
repiv posted:i don't know why we need two or more relatively slow CPU threads when the GPU thread lumps everything together and works fine, but IIRC there was pushback to a CPU megathread last time it came up shsc just loves segmentation, there's like three separate work chat threads. see klyith mentioning overclock chat when, hey, there's already an overclocking megathread!  "nuhhhh there's weird fiddly platfrom bullshit" there's nothing stopping that being discussed in a combined x86 thread. ridiculous
|
|
#
?
Oct 31, 2023 00:40
|
|
|
Farmer Crack-rear end posted:shsc just loves segmentation, there's like three separate work chat threads. see klyith mentioning overclock chat when, hey, there's already an overclocking megathread! I'd love to see a unified x86 thread, ideally with a OP that mentions Centaur and VIA.
|
|
#
?
Oct 31, 2023 00:56
|
|
|
Twerk from Home posted:I'd love to see a unified x86 thread, ideally with a OP that mentions Centaur and VIA. I won't stand for this Transmeta erasure!
|
|
#
?
Oct 31, 2023 03:11
|
|
|
Have an aunt used to live like 1 mile from the VIA hq in Fremont. That KT266A chipset was an absolute game changer back in the day. KT333 was decent too, the first to support ATA133 speeds iirc.
|
|
#
?
Nov 1, 2023 06:43
|
|
|
Intel’s RDRAM excursion and amd’s early reluctance to dabble in chipsets unless they absolutely had to left the door wide open for VIA, and they did have a short hot streak. Their only real competition was ALi, lol. Once Nvidia (and ATI, and even SiS for a bit) entered the market, it became real obvious how bad VIAs compatibility and performance really was and they faded fast out of the market. But I guess so did everyone once Intel and AMD went full anti competition and killed an entire market off forever, while the courts twiddled their thumbs.
|
|
#
?
Nov 1, 2023 09:15
|
|
|
At this point I think Centaur, VIA, and Transmeta x86 might need a retrocomputing thread
|
|
#
?
Nov 1, 2023 17:19
|
|
|
Cygni posted:Intel’s RDRAM excursion and amd’s early reluctance to dabble in chipsets unless they absolutely had to left the door wide open for VIA, and they did have a short hot streak. Their only real competition was ALi, lol. To some extent, I do miss the days of having to research a few different chipset options to see which one made the most sense for me.
|
|
#
?
Nov 1, 2023 17:31
|
|
|
FuturePastNow posted:At this point I think Centaur, VIA, and Transmeta x86 might need a retrocomputing thread What about a Zhaoxin megathread
|
|
#
?
Nov 1, 2023 17:40
|
|
|
gradenko_2000 posted:What about a Zhaoxin megathread All under the umbrella of garbage computing
|
|
#
?
Nov 1, 2023 18:00
|
|
|
drat no love for cyrix
|
|
#
?
Nov 1, 2023 18:20
|
|
|
Farmer Crack-rear end posted:drat no love for cyrix Cyrix 6x86 p166 was the first cpu I ever bought
|
|
#
?
Nov 1, 2023 23:19
|
|
|
Zen 4 + Zen4c hybrid designs officially on the market, and tries to set a new bar for bad marketing names: https://www.anandtech.com/show/21111/amd-unveils-ryzen-7040u-series-with-zen-4c-smaller-cores-bigger-efficiency quote:The first of the two new Ryzen 7040U processors with Zen 4c is the Ryzen 5 7545U, which coincidentally shares similar specifications to the existing Ryzen 7540U processor. The critical difference is that the Ryzen 5 7545U uses two full-fat Zen 4 cores with four of the smaller Zen 4c cores, which is an interesting implementation, to say the least. The second is the Ryzen 3 7440U, a direct successor to the...well, processor with the same name, the Ryzen 3 7440U, trading off most of its Zen 4 CPU cores for Zen 4c cores.  Yes, there are two SKUs with the same name... one that has 4 big cores, one that is 1 big core and 3 little cores. Also, AMD is purposefully obfuscating what the core counts are going to be on hybrid products, including launching the Z1 earlier this year that already had Zen 4c cores ... and they just didn't tell anyone. quote:Because Zen 4c has the same IPC as Zen 4, AMD is treating them as one and the same throughout their marketing and technical disclosures. And while there’s something to be said for keeping things simple for the masses, we none the less have come away from AMD’s briefing with some reservations about the matter – and some concerns about what is not being said. AMD, Intel, and Nvidia: forever locked in a war to have the worst, most consumer unfriendly marketing.
|
|
#
?
Nov 2, 2023 20:35
|
|
|
Yeah ok that is hosed up naming. They're using one core to gain their headline max turbo speed, failing to mention all the others can't come close. Bad AMD. This is going to cause a lot of noise and backlash from nerds. HalloKitty fucked around with this message at 08:00 on Nov 3, 2023 |
|
#
?
Nov 3, 2023 07:57
|
|
|
lol the 7440U is like... an ethics violation
|
|
#
?
Nov 3, 2023 08:06
|
|
|
HalloKitty posted:Yeah ok that is hosed up naming. To play devil's advocate, that's kind of how max turbo speeds end up working on everything, especially low power mobile chips. You only ever see it with 1c or 2c load, and that gets dispatched to a preferred core. On the other hand, 7000 series APU naming was already imo the worst of anyone's current offerings, with a bunch being straight up rebranding of 6000 and 5000 series chips to make them look newer. E: like the 7730U is just blatantly a 5800U with a 100mhz clock bump. What the gently caress is that. VorpalFish fucked around with this message at 15:17 on Nov 3, 2023 |
|
#
?
Nov 3, 2023 15:11
|
|
|
|
| # ? Jun 8, 2024 20:30 |
|
|
VorpalFish posted:To play devil's advocate, that's kind of how max turbo speeds end up working on everything, especially low power mobile chips. You only ever see it with 1c or 2c load, and that gets dispatched to a preferred core. Friend, have you taken a look at the Intel thread and their "14th generation" desktop chips?
|
|
#
?
Nov 3, 2023 15:16
|
|