News Nvidia enjoys $130B annual earnings despite gaming segment 'supply constraints

[Admin]

Nvidia posts record-breaking revenue, makes bold predictions amid skyrocketing demand for AI GPUs.

Nvidia enjoys $130B annual earnings despite gaming segment 'supply constraints : Read more

 

[magbarn]

Unless Ai crashes, this is why the 5090 will always be in short supply until end of production.

 

[vanadiel007]

When you fly high, you will crash hard someday. They cannot sustain these numbers.

 

[A Stoner]

Despite... Despite? No, because of the gaming card constraints. They can sell AI silicone for $60,000 and can sell graphics cards for $1000 to $2000. The AI is sold directly by themselves, and most of the $1000 and $2000 cards are sold by their AIB partners, meaning they do not even get the full sales price like they do for selling AI cards.

The restrictions are deliberate and are why they make so much money.

 

[spongiemaster]

When you fly high, you will crash hard someday. They cannot sustain these numbers.

They don't have to. They pulled in $73 billion in profits in 2024. AMD generated $1.64 billion which was twice what they did in 2023. $73 billion is likely more profits than AMD has generated in their entire 55 years as a company with all earnings adjusted for inflation. Nvidia could see a 75% decrease in profits and still be an incredibly successful company financially.

Gaming revenue for Nvidia was only 7% of their revenue in Q4. At this point, there has to be some serious thought of what Nvidia plans to do going forward. What's the point of them even staying in the market any longer? Do they spin it off and relieve themself of the headache of dealing with such a finnicky market?

 

[redgarl]

There is no constraints in the gaming segment, there is only no real motivation to do more.

The dGPU market is barelly a 6B$ TAM while the Accelerator TAM is about to hit 1T$ in 2028.

 

[redgarl]

They don't have to. They pulled in $73 billion in profits in 2024. AMD generated $1.64 billion which was twice what they did in 2023. $73 billion is likely more profits than AMD has generated in their entire 55 years as a company with all earnings adjusted for inflation. Nvidia could see a 75% decrease in profits and still be an incredibly successful company financially.

Gaming revenue for Nvidia was only 7% of their revenue in Q4. At this point, there has to be some serious thought of what Nvidia plans to do going forward. What's the point of them even staying in the market any longer? Do they spin it off and relieve themself of the headache of dealing with such a finnicky market?

The thing is this will change eventually. The market will democratize itself and other chipmakers will get a big part of the pie. It is just a matter of time. 2030 is going to be a good indicator.

 

[acadia11]

Yet the markets are giving them a bruising? AMD also took a hit after a record year, so … is it just profit taking or shoe to drop? The “battle of the bit” between the west really the US and China may be weighing heavily on access to AI technology and the inevitable clever solutions that will arise to circumvent this Maginot line. Have we reached the rubicon? I say all this to say when the .… is nvidia going to get off their collective greedy a….s and get50series GPU to the masses!

 

[acadia11]

The thing is this will change eventually. The market will democratize itself and other chipmakers will get a big part of the pie. It is just a matter of time. 2030 is going to be a good indicator.

Have you looked at the world lately, the people don’t want democracy, or rather the elephants are squashing these very ideals. Across all verticals although information flows freely it has made it easy for a very few to monopolize the usage of that information, meaning, it’s easier today to quell a competitor because it’s easier to dominate a particular information space and echo it at warp speed. With the exception of ARM which I would argue is that elephant in the RISC space although many use its tech… what tech industry is democratic when it comes to key components for cutting edge? The cost of entry is simply too high … I would argue it’s the exact opposite you’ll get consolidation like every other Industrial Revolution. And only because it’s early on can little fish play … it’s like a star you got all this stuff floating about and it eventually consolidates and coalesces around the bits with greatest gravity which ironically just makes the behemoths bigger…. And so on … inertia and gravity … don’t see this changing.

 

[hotaru251]

Jensen himself stated years ago they are no longer a graphics company they are an AI company just havent changed the name.

Gaming gpu are lost profit to them as why sell a $1200~ 90 tier gpu when can sell a multi thousand dollar ai chip?

 

[TerryLaze]

Gaming revenue for Nvidia was only 7% of their revenue in Q4. At this point, there has to be some serious thought of what Nvidia plans to do going forward. What's the point of them even staying in the market any longer? Do they spin it off and relieve themself of the headache of dealing with such a finnicky market?

If you think gaming is finnicky....jeesh
GPUs will not be the best thing for AI for ever, unless nvidia comes up with the next big thing or can at least also provide it they will go back to before AI and gaming will once again be their main income.
I mean quantum makes massive improvements and could overtake normal AI in the next years, and yes nvidia is also trying hard in quantum so they might be alright, I'm just saying.

 

[helper800]

I mean quantum makes massive improvements and could overtake normal AI in the next years...

What are you comparing with the above statement? Quantum computing and AI are two completely different things. Quantum computing is a way of solving very niche type problems and AI is whatever you want it to be. There is almost zero overlap of AI and quantum computing, with exception of AI using quantum computing to solve for some obscure problem or doing some training with it... Please explain if that is out of left field...

 

[JTWrenn]

When you fly high, you will crash hard someday. They cannot sustain these numbers.

Maybe, but it's not like they are way out over their ski's. Their p/e is about 42 and that puts them about on par with Amazon.

Some day they will lose their edge and someone else will take the lead, but right now I wouldn't bet against either of those companies.

 

[TerryLaze]

What are you comparing with the above statement? Quantum computing and AI are two completely different things. Quantum computing is a way of solving very niche type problems and AI is whatever you want it to be. There is almost zero overlap of AI and quantum computing, with exception of AI using quantum computing to solve for some obscure problem or doing some training with it... Please explain if that is out of left field...

There is no overlap now.
Because quantum is extremely weak now, the end goal of quantum is to recreate how a brain works, could your brain do what AI does now?! Yes or no?!

 

[sseemaku]

There are so many companies and startups making AI accelerators but nobody seems to be making inroads. In consumer market, people may be used to nvidia but I wonder why are AMD's MI series of GPUs not gaining any traction! All enterprises only seem to care about nvidia GPUs.

 

[JamesJones44]

Gaming revenue for Nvidia was only 7% of their revenue in Q4. At this point, there has to be some serious thought of what Nvidia plans to do going forward. What's the point of them even staying in the market any longer? Do they spin it off and relieve themself of the headache of dealing with such a finnicky market?

I don't think they would spin off the gaming unit unless margins fall and start to drag down other numbers. I also find it very unlikely they would exit the business completely given the lack of competition. What I would worry about is lack of investment in their gaming unit given how small it has become and the relatively nonexistent competition.

 

[JamesJones44]

When you fly high, you will crash hard someday. They cannot sustain these numbers.

Even if they can't, much like Apple, by the time the run slows/ends/levels off, they will have so much cash they can enter whatever market they want via acquisitions. Nvidia isn't going to go back to a small gaming GPU company without some kind of major disaster or a major shift in technology they happen to miss/ignore/can't catch up to.

 

[vanadiel007]

They don't have to. They pulled in $73 billion in profits in 2024. AMD generated $1.64 billion which was twice what they did in 2023. $73 billion is likely more profits than AMD has generated in their entire 55 years as a company with all earnings adjusted for inflation. Nvidia could see a 75% decrease in profits and still be an incredibly successful company financially.

Gaming revenue for Nvidia was only 7% of their revenue in Q4. At this point, there has to be some serious thought of what Nvidia plans to do going forward. What's the point of them even staying in the market any longer? Do they spin it off and relieve themself of the headache of dealing with such a finnicky market?

Unfortunately it does not work like that. Investors will want to see increases with each quarter. Just look at Apple and how they have to do better each quarter because investors want it that way.

 

[helper800]

There is no overlap now.
Because quantum is extremely weak now, the end goal of quantum is to recreate how a brain works, could your brain do what AI does now?! Yes or no?!

Yes, our brains do what AI does better than AI right now. The only reason an AI model can do something better than an individual now is purely brute force with the underlying computational resource. The entire goal of AI is to emulate how a humans brain solves problems and currently the best AI, with few exceptions, cannot do that better than we can. There are three levels of AI; Artificial narrow intelligence, artificial general intelligence, and artificial super intelligence. Currently we are just barely getting closer to artificial general intelligence. Let me define the three for you:

1. Artificial narrow intelligence (ANI) - Capable of performing one niche task at a time under a single domain with specific context. Examples of this level include; image recognition, speech recognition, natural language models, or natural language processing, fall under artificial narrow intelligence.

2. Artificial general intelligence (AGI) - Capable of performing multiple tasks, as a human would at a humans level of ability, across multiple relevant domains. It would be capable of learning and adapting newfound knowledge across multiple domains. A couple examples of attempts to move towards AGI, but are clearly not there, include; OpenAI, Deepmind, and many others.

3. Artificial super intelligence (ASI) - This level surpasses humans' level of cognition and ability to learn and adapt across all domains simultaneously. This level would have developed emotions, beliefs, decision-making methodologies, information processing strategies, and beyond human levels of IQ. There are no examples of even progress towards such a level.

If you want to argue with the above, talk to Mehak and Megha Kalsi, which is where I derived most of the above.