Support NeoGAF

[HeisenbergFX4]

0:00:00 Introduction
0:01:08 Blackwell architecture: Blackwell SM design
0:03:39 Updated Tensor Core and RT Core
0:09:37 AI Management Processor and Max Q power management
0:15:13 Display Engine and video encode/decode
0:18:32 Hardware specs: RTX 5080 and 5090
0:29:16 RTX 5070 and 5070 Ti
0:34:23 RTX 50 Series laptops
0:40:14 RTX Software: Neural Shaders
0:44:24 SER 2.0
0:47:06 RTX Mega Geometry
0:50:15 RTX Hair
0:53:13 Neural Radiance Cache
0:55:58 RTX Skin
0:58:01 DLSS 4: Super Resolution, Ray Reconstruction, Frame Generation
1:06:18 Generative AI demos
1:10:03 Wrap up discussion: RTX 50 series price and performance
1:17:36 Nvidia’s software package and features
1:22:38 How should we review graphics hardware?

 

[mansoor1980]

look at all those buzz words
multi-frame gen is missing from the list

 

[Imtjnotu]

I'll wait for an adamsapple breakdown tonight

 

[adamsapple]

I'll wait for an adamsapple breakdown tonight

 

[twilo99]

 

[Fess]

I'll wait for an adamsapple breakdown tonight

TLDR
It is safe to not upgrade now.

 

[Imtjnotu]

TLDR
It is safe to not upgrade now.

Thank you for not being worthless like Adam's

 

[MazingerDUDE]

So, they've been saying it's wrong to go 30 to 60 in a console game using the frame gen, and now they think it's a ok to go from 40 to 160 with now 3 times the frame gen. What complete sell outs

 

[Fess]

Thank you for not being worthless like Adam's

Joking a bit there of course. Watch the wrap up at 1:10. Oliver is more positive than Alex, think the price versus performance is surprisingly decent. Alex is more negative and think you don’t need to upgrade if you have a good 40-serie card.

Another video show that there isn’t much latency added when activating 4X FG, single digit differences per X factor. That’s the most impressive aspect of these cards imo.
But those who don’t like FG still won’t like it, total latency is around 40-60ms in CP2077, not for frame counters.

 

[Thebonehead]

AI Just took your job....
Chapter 1: Introduction

• The video introduces the Nvidia RTX 50-Series Blackwell architecture and specs.
• It discusses the new SM design, tensor cores, RT core, AI management processor, and Shader execution reordering.
• It also covers the new features such as DLSS 4, Ray Reconstruction, and Neural Texture Compression.
• Finally, the video compares the RTX 5090 and RTX 5080 and discusses the price and performance of the cards.

Chapter 2: SM Design

• This chapter discusses the new SM design in the RTX 50-Series Blackwell architecture.
• The new SM design has 32 streaming multiprocessors (SMs), each with 1024 CUDA cores.
• This is a significant increase from the previous generation of SMs, which had 64 SMs with 2048 CUDA cores.
• The new SM design also features a new AI management processor (AMP), which is a dedicated AI accelerator.

Chapter 3: Tensor Cores

• This chapter discusses the new tensor cores in the RTX 50-Series Blackwell architecture.
• The new tensor cores are 4x faster than the previous generation of tensor cores.
• They are also more efficient, consuming less power.
• Tensor cores are used for a variety of AI tasks, such as image recognition, speech recognition, and natural language processing.

Chapter 4: RT Core

• This chapter discusses the new RT core in the RTX 50-Series Blackwell architecture.
• The new RT core is 2x faster than the previous generation of RT cores.
• It is also more efficient, consuming less power.
• RT cores are used for ray tracing, which is a technique for rendering realistic images and videos.

Chapter 5: AI Management Processor

• This chapter discusses the new AI management processor (AMP) in the RTX 50-Series Blackwell architecture.
• The AMP is a dedicated AI accelerator that is used to accelerate a variety of AI tasks.
• It is also used to improve the performance of DLSS 4.

Chapter 6: Shader Execution Reordering

• This chapter discusses the new Shader execution reordering in the RTX 50-Series Blackwell architecture.
• Shader execution reordering is a technique that can improve the performance of graphics applications.
• It allows the GPU to execute instructions in a different order than they were originally written.

Chapter 7: DLSS 4

• This chapter discusses the new DLSS 4 feature in the RTX 50-Series Blackwell architecture.
• DLSS 4 is a super-resolution technology that can increase the frame rate of a game without sacrificing image quality.

Chapter 8: Ray Reconstruction

• This chapter discusses the new Ray Reconstruction feature in the RTX 50-Series Blackwell architecture.
• Ray Reconstruction is a technique that can improve the quality of ray-traced images.

Chapter 9: Neural Texture Compression

• This chapter discusses the new Neural Texture Compression feature in the RTX 50-Series Blackwell architecture.
• Neural Texture Compression is a technique that can reduce the size of textures without sacrificing image quality.

Chapter 10: RTX 5090 vs RTX 5080

• This chapter compares the RTX 5090 and RTX 5080.
• The RTX 5090 is the fastest and most expensive of the two cards.
• The RTX 5080 is a more affordable option that still offers excellent performance.

Chapter 11: Price and Performance

• This chapter discusses the price and performance of the RTX 50-Series Blackwell architecture.
• The RTX 50-Series cards are more expensive than the previous generation of RTX cards.
• However, they also offer significantly better performance.

 

[JaksGhost]

AI Just took your job....
Chapter 1: Introduction

• The video introduces the Nvidia RTX 50-Series Blackwell architecture and specs.
• It discusses the new SM design, tensor cores, RT core, AI management processor, and Shader execution reordering.
• It also covers the new features such as DLSS 4, Ray Reconstruction, and Neural Texture Compression.
• Finally, the video compares the RTX 5090 and RTX 5080 and discusses the price and performance of the cards.

Chapter 2: SM Design

• This chapter discusses the new SM design in the RTX 50-Series Blackwell architecture.
• The new SM design has 32 streaming multiprocessors (SMs), each with 1024 CUDA cores.
• This is a significant increase from the previous generation of SMs, which had 64 SMs with 2048 CUDA cores.
• The new SM design also features a new AI management processor (AMP), which is a dedicated AI accelerator.

Chapter 3: Tensor Cores

• This chapter discusses the new tensor cores in the RTX 50-Series Blackwell architecture.
• The new tensor cores are 4x faster than the previous generation of tensor cores.
• They are also more efficient, consuming less power.
• Tensor cores are used for a variety of AI tasks, such as image recognition, speech recognition, and natural language processing.

Chapter 4: RT Core

• This chapter discusses the new RT core in the RTX 50-Series Blackwell architecture.
• The new RT core is 2x faster than the previous generation of RT cores.
• It is also more efficient, consuming less power.
• RT cores are used for ray tracing, which is a technique for rendering realistic images and videos.

Chapter 5: AI Management Processor

• This chapter discusses the new AI management processor (AMP) in the RTX 50-Series Blackwell architecture.
• The AMP is a dedicated AI accelerator that is used to accelerate a variety of AI tasks.
• It is also used to improve the performance of DLSS 4.

Chapter 6: Shader Execution Reordering

• This chapter discusses the new Shader execution reordering in the RTX 50-Series Blackwell architecture.
• Shader execution reordering is a technique that can improve the performance of graphics applications.
• It allows the GPU to execute instructions in a different order than they were originally written.

Chapter 7: DLSS 4

• This chapter discusses the new DLSS 4 feature in the RTX 50-Series Blackwell architecture.
• DLSS 4 is a super-resolution technology that can increase the frame rate of a game without sacrificing image quality.

Chapter 8: Ray Reconstruction

• This chapter discusses the new Ray Reconstruction feature in the RTX 50-Series Blackwell architecture.
• Ray Reconstruction is a technique that can improve the quality of ray-traced images.

Chapter 9: Neural Texture Compression

• This chapter discusses the new Neural Texture Compression feature in the RTX 50-Series Blackwell architecture.
• Neural Texture Compression is a technique that can reduce the size of textures without sacrificing image quality.

Chapter 10: RTX 5090 vs RTX 5080

• This chapter compares the RTX 5090 and RTX 5080.
• The RTX 5090 is the fastest and most expensive of the two cards.
• The RTX 5080 is a more affordable option that still offers excellent performance.

Chapter 11: Price and Performance

• This chapter discusses the price and performance of the RTX 50-Series Blackwell architecture.
• The RTX 50-Series cards are more expensive than the previous generation of RTX cards.
• However, they also offer significantly better performance.

I might crack on him from time-to-time but Adam gives us the fine details in his breakdowns which matters. This is trash.

 

[llien]

Everyone and their grandma is shamelessly citing Computerbase. (these guys do rock)