Apple built a DirectX 12 translation layer for Apple Silicon

[koyoot]

I was referring to the total gaming industry.

Compared to the camera industry.

Based on this, do you think it's more likely that Apple makes a gaming console or a standalone camera?

This stat includes MOBILE gaming. On smartphones and tablets.

 

[jmho]

This discussion always reminds me of the fact that when I was a kid, if you wanted your computer to do floating point numbers, you'd need an expensive FPU add-on card.

Eventually cheap CPUs started adding slow FPUs, which were no where near as fast as the stand-alone FPUs you'd find in high end systems.

I'm sure the stand-alone FPU people mocked the "integrated FPU" noobs with their dumb casual FPUs that could never compete with a "real" FPU.

 

[Xiao_Xi]

Based on leaked GFX Bench scores, and this is graphics workload, NOT COMPUTE, M2 Ultra is 25% faster than RTX 4090 desktop GPU.

On which benchmarks? The website says otherwise.

GFXBench - Unified cross-platform 3D graphics benchmark database

The first unified cross-platform 3D graphics benchmark database for comparing Android, iOS, Windows 8, Windows Phone 8 and Windows RT capable devices based on graphics processing power.

gfxbench.com

 

[Longplays]

I was referring to the total gaming industry.

Edit: To add emphasis as it was pointed out that the graph includes smartphone/tablet gaming

Red: Social/casual gaming (smartphone/tablets)

Yellow: PC gaming

Orange: game console

.... Seeming apple makes more money than almost everyone via App Store... why be agressive with gaming console?

Compared to the camera industry.

Based on this, do you think it's more likely that Apple makes a gaming console or a standalone camera?

It is likely Apple will improve Mac gaming.

Dedicated gaming console at par is very unlikely as it would force Apple to follow the business model of Xbox/PS.

Next Apple TV 4K or even an 8K after 2026 may be a game console.

 

[koyoot]

On which benchmarks? The website says otherwise.

GFXBench - Unified cross-platform 3D graphics benchmark database

The first unified cross-platform 3D graphics benchmark database for comparing Android, iOS, Windows 8, Windows Phone 8 and Windows RT capable devices based on graphics processing power.

gfxbench.com

In the quoted by you post I have linked sites with benchmarks, and DIRECT COMPARISONS to other Laptop GPUs.

Notebookcheck not only gives you scores for Mac GPUs but also compares them directly to their competitors.

You didn't bother with reading those sites?

 

[diamond.g]

In the quoted by you post I have linked sites with benchmarks, and DIRECT COMPARISONS to other Laptop GPUs.

Notebookcheck not only gives you scores for Mac GPUs but also compares them directly to their competitors.

You didn't bother with reading those sites?

They quoted you saying 4090 Desktop GPU, but you meant Laptop GPU?

 

[Xiao_Xi]

In the quoted by you post I have linked sites with benchmarks, and DIRECT COMPARISONS to other Laptop GPUs.

I was referring to the link that shows that M2 Ultra is 25% faster than RTX 4090 on GFXBench. GFXBench has many benchmarks, so it could be true for one of them.

M2 Ultra is 25% faster than RTX 4090 desktop GPU.

 

[senttoschool]

This discussion always reminds me of the fact that when I was a kid, if you wanted your computer to do floating point numbers, you'd need an expensive FPU add-on card.

Eventually cheap CPUs started adding slow FPUs, which were no where near as fast as the stand-alone FPUs you'd find in high end systems.

I'm sure the stand-alone FPU people mocked the "integrated FPU" noobs with their dumb casual FPUs that could never compete with a "real" FPU.

But FPUs are once again going discrete.

 

[jeanlain]

In the quoted by you post I have linked sites with benchmarks, and DIRECT COMPARISONS to other Laptop GPUs.

Notebookcheck not only gives you scores for Mac GPUs but also compares them directly to their competitors.

You didn't bother with reading those sites?

Where is it shown that the M2 ultra is 25% faster than the desktop 4090, neither being laptop GPUs?
The GFXBench database shows the opposite. https://gfxbench.com/compare.jsp?be...me1=Apple+M2+Ultra&D2=NVIDIA+GeForce+RTX+4090

 

[senttoschool]

Where is it shown that the M2 ultra is 25% faster than the desktop 4090, neither being laptop GPUs?
The GFXBench database shows the opposite. https://gfxbench.com/compare.jsp?be...me1=Apple+M2+Ultra&D2=NVIDIA+GeForce+RTX+4090

Weird numbers. It appears that M2 Max is faster in some benchmarks.

GFXBench - Unified cross-platform 3D graphics benchmark database

The first unified cross-platform 3D graphics benchmark database for comparing Android, iOS, Windows 8, Windows Phone 8 and Windows RT capable devices based on graphics processing power.

gfxbench.com

 

[jeanlain]

Weird numbers. It appears that M2 Max is faster in some benchmarks.

GFXBench - Unified cross-platform 3D graphics benchmark database

The first unified cross-platform 3D graphics benchmark database for comparing Android, iOS, Windows 8, Windows Phone 8 and Windows RT capable devices based on graphics processing power.

gfxbench.com

Something's fishy. We need more results from the M2 ultra.

 

[diamond.g]

Something's fishy. We need more results from the M2 ultra.

I wish Apple would give UL Solutions a giant bag of money to convert 3dmark test to their Metal equivalent (aside from Wildlife). Then we can stop using GFXBench, lol.

 

[Longplays]

I wish Apple would give UL Solutions a giant bag of money to convert 3dmark test to their Metal equivalent (aside from Wildlife). Then we can stop using GFXBench, lol.

People would cry "bias" as money changed hands.

 

[diamond.g]

People would cry "bias" as money changed hands.

I am sure MS pays UL money to make DirectX benchmarks. Why else would they have made Speedway when Port Royal does basically the same thing (at the same resolution as well).

 

[koyoot]

I was referring to the link that shows that M2 Ultra is 25% faster than RTX 4090 on GFXBench. GFXBench has many benchmarks, so it could be true for one of them.

Where is it shown that the M2 ultra is 25% faster than the desktop 4090, neither being laptop GPUs?
The GFXBench database shows the opposite. https://gfxbench.com/compare.jsp?be...me1=Apple+M2+Ultra&D2=NVIDIA+GeForce+RTX+4090

Here. 317 FPS. Aztec Ruins 4K OffScreen, 317 FPS and that score is compared to the scores on Notebookcheck database with other laptops.

They quoted you saying 4090 Desktop GPU, but you meant Laptop GPU?

Yes, actually - its 4090 laptop GPU, which is based on full AD103 die. And M2 Ultra is in this particular benchmark 25% faster: 242 FPS for 4090 laptop, and 317 FPS for the M2 Ultra 76 core GPU.

 

[diamond.g]

Here. 317 FPS. Aztec Ruins 4K OffScreen, 317 FPS and that score is compared to the scores on Notebookcheck database with other laptops.

Yes, actually - its 4090 laptop GPU, which is based on full AD103 die. And M2 Ultra is in this particular benchmark 25% faster: 242 FPS for 4090 laptop, and 317 FPS for the M2 Ultra 76 core GPU.

Right the Laptop 4090 is a lower frequency Desktop 4080. The actual Desktop 4090 is like another 25% faster than the Desktop 4080.

 

[dmccloud]

I wish Apple would give UL Solutions a giant bag of money to convert 3dmark test to their Metal equivalent (aside from Wildlife). Then we can stop using GFXBench, lol.

The last thing the industry needs is another benchmark scandal, especially when nVidia hasn't learned from its past mistakes:



https://www.gamespot.com/articles/nvidia-accused-of-cheating-in-3dmark-03/1100-6028894/

Nvidia accused of cheating in big-data performance test by benchmark's umpires: Workloads 'tweaked' to beat rivals in TPCx-BB

GPU giant says it'll play ball soon

www.theregister.com

 

[Geminiwave]

anybody see how the vanilla M2 performs? I'm really impressed with the Max and Ultra results, buuuuut I'm wondering if my purchase of an M2 Air was a poor one. Back when gaming wasn't really on the table, the M2 seemed fine.

 

[TSE]

Does anyone know if this works for older games that run off DirectX 9?

 

[dmccloud]

anybody see how the vanilla M2 performs? I'm really impressed with the Max and Ultra results, buuuuut I'm wondering if my purchase of an M2 Air was a poor one. Back when gaming wasn't really on the table, the M2 seemed fine.

I've seen multiple people running some of these titles on the base M1 Air, so the M2 should have at least some level of performance above that.

 

[Longplays]

Does anyone know if this works for older games that run off DirectX 9?

I hope it runs DirectX 1-9 games!

 

[Numa_Numa_eh]

I hope it runs DirectX 1-9 games!

Game Porting Kit is Direct X 11 and 12. Crossover already does 1-9

 

[dmr727]

I'm just happy I'll be able to play Skyrim on yet another platform without having to give those vampires at Bethesda another dollar.

 

[senttoschool]

anybody see how the vanilla M2 performs? I'm really impressed with the Max and Ultra results, buuuuut I'm wondering if my purchase of an M2 Air was a poor one. Back when gaming wasn't really on the table, the M2 seemed fine.

I've seen some M2 Air 8GB running games and it's surprisingly good given the low RAM, no fan, Rosetta2 translation, and DX12 translation.

 

[diamond.g]

The last thing the industry needs is another benchmark scandal, especially when nVidia hasn't learned from its past mistakes:



https://www.gamespot.com/articles/nvidia-accused-of-cheating-in-3dmark-03/1100-6028894/

Nvidia accused of cheating in big-data performance test by benchmark's umpires: Workloads 'tweaked' to beat rivals in TPCx-BB

GPU giant says it'll play ball soon

www.theregister.com

As long as Apple doesn't cheat like nvidia did (I remember that scandal, lol) it'll be fine.

Does anyone know if this works for older games that run off DirectX 9?

I hope it runs DirectX 1-9 games!

Game Porting Kit is Direct X 11 and 12. Crossover already does 1-9

So I recall seeing someone say that this converter supports DX9on12 emulation (which is MS official way to run older games on newer hardware; just ask Intel, lol) but I am unsure if that extends to older versions of DX.