Watch Nvidia's GTC 2020 Keynote here

While watching this Keynote I realized that while AMD CPU division is kicking Intel's behind AMD Radeon Group, with their few puny cards, is so insignificant compared to Nvidia machinery. The whole World is running on Nvidia hardware and AMD has nothing even close neither on the hardware nor on the software level.
 
While watching this Keynote I realized that while AMD CPU division is kicking Intel's behind AMD Radeon Group, with their few puny cards, is so insignificant compared to Nvidia machinery. The whole World is running on Nvidia hardware and AMD has nothing even close neither on the hardware nor on the software level.

It's easy to see Nvidia as a gaming company, but it just isn't anymore. When it comes to AI, Nvidia is running rings around Intel, and AMD isn't even in contention for a lot of the markets that Nvidia is in. There is a reason why they have grown so much in recent years.

AMD fell far behind Nvidia, but hopefully things are on the way back up for them with RDNA 2. It seems to be a big architecture for them, and we will have to see how that goes for them. A 50% performance/watt bump is huge, especially without a new node. The question no is if it is enough.

As far as Nvidia goes, they are doubling down on AI and will start to push DLSS harder. Even if RDNA 2 is great, I don't think AMD has an answer to DLSS. That will be a big deal for them.

As you said, Nvidia is on a different scale to AMD in many ways, just like how Intel was with AMD. I just want to see AMD succeed, not because I am an AMD fan, but because we need competition to keep Nvidia's prices down and to keep engineers pushing every market forward. I also want to see Intel succeed in the GPU market for the same reason.

I think it became very clear today why Nvidia purchased Mellanox for $6.9 billion. They aren't a gaming company anymore, they are a gaming and datacenter company.
 
To be fair all the big tech companies were never about consumers. They used consumers to fund R&D for the bigger markets. That was and always has been the plan. That is where the high margin products sell. There is a reason the fastest growing markets for these companies is in higher-margin markets. Also kinda the reason Amazon launched the AWS Cloud business and launched cloud business to the world and is easily on pace for a $1 trillion market by 2030.

AMD is an exception here somewhat as they are growing in all markets but that is due to them being so nonexistent in any to start with.
 
It's easy to see Nvidia as a gaming company, but it just isn't anymore. When it comes to AI, Nvidia is running rings around Intel, and AMD isn't even in contention for a lot of the markets that Nvidia is in. There is a reason why they have grown so much in recent years.

AMD fell far behind Nvidia, but hopefully things are on the way back up for them with RDNA 2. It seems to be a big architecture for them, and we will have to see how that goes for them. A 50% performance/watt bump is huge, especially without a new node. The question no is if it is enough.

As far as Nvidia goes, they are doubling down on AI and will start to push DLSS harder. Even if RDNA 2 is great, I don't think AMD has an answer to DLSS. That will be a big deal for them.

As you said, Nvidia is on a different scale to AMD in many ways, just like how Intel was with AMD. I just want to see AMD succeed, not because I am an AMD fan, but because we need competition to keep Nvidia's prices down and to keep engineers pushing every market forward. I also want to see Intel succeed in the GPU market for the same reason.

I think it became very clear today why Nvidia purchased Mellanox for $6.9 billion. They aren't a gaming company anymore, they are a gaming and datacenter company.

Even if RDNA 2 is a miracle that AMD needs I can only see them competing up to the high-medium level Gaming GPU segment. They probably won't have a 3080 Ti killer, or anything near it. That may bring prices a bit down, but AMD might want a piece of that expensive cookie that Nvidia has baked for the customers.

Datacenter, server, and AI markets are just impossible for them.

Professional segment (Architecture, Animation, 3D modeling, Video Editing/Rendering, even Streaming) is also out of AMD's reach. Every single rendering engine (Cycles, V-Ray, Octane, ...) all of them have CUDA and RTX specific optimizations. There is no way that AMD can produce a GPU that can beat Nvidia in applications that are tailored for and co-developed by the green team.

The only hope is that AMD can just nudge Nvidia a bit in the Gaming segment so they don't become Intel from the past 10 years.
 
Back
Top