Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
My M1 is 65Watts, pc 650 watts.
That is one of the upsides, while my M4 Max Studio consumes more power then your M1, its significantly less then my PC. When playing cyperpunk on my PC, the PC and GPU fans are ramping up and while not obscenely loud, its louder then I prefer. While I posted the differencs in FPS between my PC and Mac desktop, I will say the playing experience on my Mac is better partly due to the quietness of my Mac. Plus (especially in the summer), I'm not heating up my room as I play games, which is the case with my PC
 
That is one of the upsides, while my M4 Max Studio consumes more power then your M1, its significantly less then my PC. When playing cyperpunk on my PC, the PC and GPU fans are ramping up and while not obscenely loud, its louder then I prefer. While I posted the differencs in FPS between my PC and Mac desktop, I will say the playing experience on my Mac is better partly due to the quietness of my Mac. Plus (especially in the summer), I'm not heating up my room as I play games, which is the case with my PC

This is partly why I don’t use my PS5 a lot during the summer, because my OLED TV can kick out some heat and it just heats an already hot room up. My Claw 8 and MacBook Pro are nothing like that.
 
DLSS doing heavy lifting for sure.
In handheld it goes all the way down to 360p to target 40fps.

I don't have the game but if I lowered the resolution to 540p or 360p and capped the FPS at 40 the power draw would also be quite low on any Apple Silicon machine.
 
In handheld it goes all the way down to 360p to target 40fps.

I don't have the game but if I lowered the resolution to 540p or 360p and capped the FPS at 40 the power draw would also be quite low on any Apple Silicon machine.
Would it hit 40FPS at 10W? Or even 30? I'm skeptical even if you used an external monitor to get the screen out of the power draw equation...

Steam Deck, which can also be very efficient gets about 20 FPS on the absolute lowest settings at 720p with XeSS Ultra Performance (so 360p rendering resolution) when limited to 10W. And it looks like a total potato.
 
Given that the game is playable in my M1 with 16GB ram. I am confident with improvements to metal fx, iphone models in 1-2 years should be able to match the Switch 2.

The big barrier on iPhones is space. Apple need to increase base storage to 256GB so that 1-2 big games can be installed without anxiety about storage.
 
Given that the game is playable in my M1 with 16GB ram. I am confident with improvements to metal fx, iphone models in 1-2 years should be able to match the Switch 2.

The big barrier on iPhones is space. Apple need to increase base storage to 256GB so that 1-2 big games can be installed without anxiety about storage.
A18 Pro with a 6-core GPU is already faster than the Nintendo Switch 2 in handheld (2.46TFlops versus 1.71TFlops).

You can strip the game from 4K assets, that will drastically reduce file size.
 
  • Like
Reactions: Macalicious2011
A PC gamer trying to play Cyberpunk on a M4 Mini. Results are not surprising, but I will say that while most of us wouldn't turn off MetalFX, I see his point on why he did, i.e., for a base line on actual game play w/o any upscaling. With that said, I think its safe to assume 99% of the players will have metalfx turned on for obvious reasons

His conclusion aligns with mine - you need a pro level chip at a minimum to get any sort of acceptable game play. That doesn't mean the game won't work on anything less, but performance, say on a MBA wouldn't be ideal.

 
better results with turning RT off. I saw that @Homy is participating in this thread as well. Not sure if they set SSR to match the PC setting (I don't recall seeing it mentioned).
From my googling, reading reviews, and information about the M4 Max, its generally accepted that M4 Max is equivalent to that of a RTX 4070 so those numbers are not surprising at all. Likewise with the Ryzen 9 AI 370, in doing a quick google that CPU is equivalent to the non-pro M4.

Seems like someone wanted to try to justify apple silicon's superiority and preach to the apple fans on reeddit by using slower CPU/GPUs

I mean, I wonder how a thread on reddit would go if I stated my RX7800 XT beat out the RTX 4060 cards with RT turned off? Same sort of thing, the 4060 is a lower tiered gpu and the RX7800 is similar to a 4070.
 
  • Like
Reactions: Huntn
With that said, I think its safe to assume 99% of the players will have metalfx turned on for obvious reasons
It’s no different on PC. Unless you are playing at 1080p on an overpowered PC, you will use dlss/fsr. On my upper midrange PC, ray and path tracing are COMPLETELY unplayable without upscaling. In fact, before Nvidia launched dlss transformers this February I didn’t use ray or path tracing because even with cnn dlss the fps hit was too big.

Going forward, improvements of metalfx, will be more important than more powerful hardware.
 
Last edited:
  • Like
Reactions: Irishman and Huntn
It’s no different on PC. Unless you are playing at 1080p on an overpowered PC, you will use dlss/fsr. On my upper midrange PC, ray and path tracing are COMPLETELY unplayable without upscaling. In fact, before Nvidia launched dlss transformers this February I didn’t use ray or path tracing because even with cnn dlss the fps hit was too big.

Going forward, improvements of metalfx, will be more important than more powerful hardware.
Yeah, with no upscaling or frame gen, maxed settings, 4k, all ray tracing on avg 33.12fps. Min 28.68, max 39.77. That's on a 9800x3d/rtx5090. lol
 
  • Wow
Reactions: Huntn
It’s no different on PC. Unless you are playing at 1080p on an overpowered PC, you will use dlss/fsr. On my upper midrange PC, ray and path tracing are COMPLETELY unplayable without upscaling.
On my PC (AMD 3700X/RX 7800XT) 1080P, high settings, no RT and no FSR. I think testing how I normally play on the PC w/o RT and the Mac w/o RT is comparing apples to apples (the mac is at 1080P). I don't need to play with FSR turned on given the high frame rates, and I avoid so, since I don't have to deal with the downsides of framegen.

PC an average of 120FPS
1753439739056.png



On my M4 Max, same settings, I'm getting 60 FPS
1753439897558.png
 
  • Sad
Reactions: Irishman
On my PC (AMD 3700X/RX 7800XT) 1080P, high settings, no RT and no FSR. I think testing how I normally play on the PC w/o RT and the Mac w/o RT is comparing apples to apples (the mac is at 1080P). I don't need to play with FSR turned on given the high frame rates, and I avoid so, since I don't have to deal with the downsides of framegen.

PC an average of 120FPS
View attachment 2531813


On my M4 Max, same settings, I'm getting 60 FPS
View attachment 2531815
Wow, the Mac is all over the place. A low of 30 FPS and a high of 190? That's not great. How is it running? Mostly good with giant dips?
 
  • Sad
Reactions: Irishman
Wow, the Mac is all over the place. A low of 30 FPS and a high of 190? That's not great. How is it running? Mostly good with giant dips?
Here's with metalfx turned on and its more consistent, this is how I play and it runs really well on the mac. I'm actually quite impressed.
1753440586511.png



Turning off vsync (I tend to leave it on) causes a higher degree of variance between the lows and highs
1753440802107.png
 
  • Like
Reactions: Plutonius
On my PC (AMD 3700X/RX 7800XT) 1080P, high settings, no RT and no FSR. I think testing how I normally play on the PC w/o RT and the Mac w/o RT is comparing apples to apples (the mac is at 1080P). I don't need to play with FSR turned on given the high frame rates, and I avoid so, since I don't have to deal with the downsides of framegen.

PC an average of 120FPS
View attachment 2531813


On my M4 Max, same settings, I'm getting 60 FPS
View attachment 2531815
Apple silicone, just like the Switch 2 will going forwards, have to rely more on AI upscaling. This is fine as it is not neccessary for it to meet the rasterised performance of apowerful GPU like yours.

On my PC I always have DLSS switched on except for front simulator for which it makes the text in the cockpit a tiny bit softer. Other than that, when AI upscaling is great I have no reason not to use it.
 
I have no reason not to use it.
I can see upscaling having a net benefit, but I'm not all that jazzed up for frame generation. As I mentioned, if I'm getting performance w/o DLSS/FSR, I see no reason to have it turned on. Cyberpunk is a good example, I'm getting 120FPS on average

For the Mac it smooths things out and makes the game play a better experience.
 
  • Like
Reactions: Irishman
I can see upscaling having a net benefit, but I'm not all that jazzed up for frame generation. As I mentioned, if I'm getting performance w/o DLSS/FSR, I see no reason to have it turned on. Cyberpunk is a good example, I'm getting 120FPS on average

For the Mac it smooths things out and makes the game play a better experience.
In PC land, typically if you have wild variations in framerate it is usually seen as a CPU bottleneck (or you are running out of frame buffer). I wonder if that is the case on Mac as well.
 
  • Like
Reactions: Irishman
CPU bottleneck (or you are running out of frame buffer). I wonder if that is the case on Mac as well.
I'd be surprised and disappointed of that is the case. My PC is a 3700x and as you can see the M4 Max is destroying that processor in synthetic benchmarks, Plus with the CPU and GPU on the same piece of silicon, shouldn't these sort of bottlenecks not be an issue?

1753451025140.png
 
  • Like
Reactions: Huntn and Plutonius
I'd be surprised and disappointed of that is the case. My PC is a 3700x and as you can see the M4 Max is destroying that processor in synthetic benchmarks, Plus with the CPU and GPU on the same piece of silicon, shouldn't these sort of bottlenecks not be an issue?

View attachment 2531863

There will always be bottlenecks regardless of the CPU and GPU being used. The question is whether it is a CPU or a GPU bottleneck. There are numerous factors that affect overall performance including PCIE bandwidth, GPU VRAM, frame buffers, and memory bandwidth. I would recommend running 3D Mark (now that there is a legitimate Mac version), as that will give a better comparison in terms of gaming performance than either GeekBench or Cinebench can offer. The Port Royal benchmark is better to use when comparing performance with RT enabled.
 
  • Like
Reactions: Huntn and Irishman
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.