Home
Creators
Search Recent Random
Posts
Search DMs Popular Hash Lookup Tags Random
Importer
Import FAQ
Register Favorites Keys Review DMs Logout
Community
Telegram ThePornDude
Home Creators Posts Import Register Favorites Logout

Discussing XBOX's Future, PS5 Pro, Switch 2, and GPUs w/ NXGamer (Patreon)

Published:
2024-02-17 03:39:03
Edited:
2024-02-17 03:43:17
Imported:
2026-06
Tags:

Content

The next episode of Broken Silicon will see us be joined by NXGamer once again!  Michael is somebody I have known for years, and without going into too much detail - he is a professional in this industry with legitimate expertise in how the semiconductor industry works.  He truly is a wealth of knowledge who is more than capable to speak about PC Hardware, console hardware, or even the Video Games Industry at the macro level.  Please do not hesitate to ask us about anything you normally would from our "typical subjects" of discussion on Broken Silicon, nor should you hesitate to ask tons of questions involving PlayStation, XBOX, and Nintendo...

You have ~18 hours (Till Saturday Evening US Central Time), to submit below!!!  Be respectful, as concise as possible, and use good grammar to be considered. :)

Check out NXGamer: https://www.youtube.com/@NXGamer

Last Episode Michael was on: https://youtu.be/iFwtE2fh-y0?si=xum17Iz4QdBSJHDu0

https://youtu.be/KGlD9SO3rKU?si=-p5k4VRfZLA2Fs1E

https://youtu.be/A0s0Usfg__g?si=O7LQXUXyFxsVvrF7

https://youtu.be/Hbx4AUcQ5do?si=JiC_4bXmxNkVfGNc

https://youtu.be/ZpjBHwfIyXs?si=Ak9ZEn0pQ8bxAXrJ

Comments

Lo_Res_Gamer

Hi Tom and Michael, Taking another opportunity to flip conventional wisdom on its head. Current thought for consumer AI is that computers that don't have dedicated AI would fall back to the GPU for AI processing. Could this work in reverse as well? For example, could an AMD next gen CPU with integrated graphics leverage the NPU to boost graphical performance? What are the trade-offs?

Dr Deep

What can we except software wise from the Switch 2? Will it use the same accounts, can we expect backwards compatibility with upscaling? I would really love it if we could play better versions of our existing libraries, but knowing Nintendo they will probably double dip on those.

Mr. Tiger

I was reading a recent article on how Intel was the first to get ASML’s new High-NA Extreme UV System, and that they anticipate to be producing chips from that machine in late 2025. I find the effort behind Intel’s CEO to secure the rights to be the first to get ASML’s latest lithography machine (I read this part last year) somewhat desperate given how badly things seem to be going for them lately. So my question is can Intel make up the difference just by buying the latest lithography machine before their competition such as TSMC? Is this really a smart use of their money right now?

Jamwap

I've heard rumors that the Switch 2 will have 12gb of LPDDR5. Do you think that will actually make it more appealing to game devs than the Series S which only has 8gb of VRAM? Maybe even the Series X which only has 10gb of VRAM. Also how do you think the speed of LPDDR5 will stack up?

Dave Scholze

Hey guys, I've heard some rumors about an XBox handheld. Do you think Microsoft would be able to pull of its own "Steam Deck"? If so, what do you think the specs would be? Maybe "just" Strix Halo oder a custom APU?

10GHz Netburst

Is it just me, is it just the games I play, or is DLSS/FSR/DLAA just kinda crap? Basically every time I've tried any of them it ends up either smudgy, full of ghosting, or both, to the point that it's a noticeable downgrade to visibility and detail.

OmegaDoge

Hi Tom and Michael! Bang up job on the RDNA 4 leaks, and it helped me make the decision to buy an RX 7900 XT for US$740 after taxes, skip RDNA 4, and potentially upgrade to RDNA 5 high end when it comes out. Speaking of RDNA 4, I've been thinking that if Navi 48 will be competing with the RTX 5070, wouldn't it be better to call it the 8700 XT instead of the 8800 XT if Radeon wants to improve their mind share? What do you think? Thanks as always, and appreciate your work!

Eric Bagely

Despite the doom and gloom surrounding the discussion of Ai use in videogame development, I'm actually excited about it. It looks like many of the tools being developed will allow smaller studios to make a game that punches far above their weight class. Any insight would be appreciated.

Eric Bagely

Definition of "skin walker studio" = A formerly great studio that has made nothing but mediocre games for the past decade, coasting by on past successes. Give your 3 studios that you believe fit this definition.

Deadeyes

Where are you getting those VRAM stats for the Xbox consoles? They're not accurate. S has 10 total and X has 16 total. Some is consumed by OS and they both have split speeds but Series X is not locked down to 10GB for games.

Cleansweep

Hey guys, This very interesting article on the economics of the game industry (https://www.matthewball.vc/all/gaming2024) got me thinking: Are a lot of the economic problems the game industry's experiencing just the result of video games hitting the kind of mainstream adoption that movies and TV hit decades earlier? Maybe it's time to accept that overall audience growth is going to slow down and everything needs to be designed around long-term sustainability?

Jeffrey Reyes

Do you think NVIDIA could provide a GPU for the PS6 or next generation XBOX? Why has AMD owned this position here the last 2 console generations and could NVIDIA compete again here?

KingKoro

Hello Tom and Michael! Tom, recently you leaked that RDNA4 will have FP8 support and better Matrix Cores. This suggests a much deeper focus on AI performance for RDNA4 compared to previous RDNA-gens. In my opinion this actually necessitates a 32GB Version of the RX 8800XT, rather than making it just a consideration, since AI Devs really need as much VRAM as possible. I think an affordable (500-600$), decently efficient and compact GPU with 32GB could finally threaten NVIDIA's dominance in this field, while a 48/64GB Pro version would make sense as well. Do you have any updated info on a 32GB RDNA4 card?

JonWickBoomStik

Hi Tom with the changes in the gaming industry over the years between consoles and handheld market. Do you or Mike feel that the handheld market is making a definite resurgence in the industry and will all platforms form their own versions of handhelds to capture a portion of the market like steam has?

Dark Side of the Force

To piggy back this question... I know the steamdeck APU was of Microsoft abandonment... as you said before. However, as of recent, I've seen mixed signs between Microsoft commitment with AMD (at the AMD AI event vs Xbox looking out for other partners). How committed Microsoft to AMD is? Do AMD feel burnt by Microsoft as they have dumped them multiple time with multiple different departments? What is the state of co-op between them with the current mixed signs I mentioned above?

Athena Azuraea

I personally don't think we'll see it happening. NPUs are great for matrix and vector multiplications, as commonly found in neural algorithms (what most people like to call AI). But they completely lacks other things such as shader support, so that would mean specializing the pipeline to schedule resources between the NPU, the CPU and the GPU. Of course, it will happen. But given how big a topic "AI" is, I think it's more likely game engines will focus on offloading neural processing to the NPU rather than dealing with all kinds of other graphics processing tasks.

Donkey Kong Odyssey

For a Steam Deck/Wii U-like PlayStation handheld to be successful, what kind of hardware configuration would you believe to be in Sony's best interest? Would a separate SoC optimized for handheld use be better or would a streamlined, cut-down (and downclocked) SoC be better for yields?

Lo_Res_Gamer

I agree to some extent, with barriers to implementing it: matrix sparsity differences between the two applications and the NPU could be designed with a looser tolerance for latency than would work for graphics. OTOH, the existing nVidia frame generation is using silicon more geared to AI than to standard GPU pipelines, so I don't think we can rule it out for things like post-processing.