NVIDIA Ada Lovelace

[emil vasilev]

NVIDIA подготвя видеокарти ,които да избягват американските рестрикции за Китай.

GPU	HGX H20	L20 PCle	L2 PCle
Architecture | GPU	Hopper | GH100	Ada Lovelace | AD102	Ada Lovelace | AD104
Memory	96 GB HBM3	48 GB GDDR6 w/ ECC	24 GB GDDR6 w/ ECC
Total Processing Power (FP16/BF16)	2,368	1,912	1,544
Performance Density	2.9	3.13	5.2
Memory Bandwidth	4.0 TB/s	864 GB/s	300 GB/s
INT8 I FP8 Tensor	296 I 296 TFLOPS	239 I 239 TFLOPS	193 I 193 TFLOPS
BF16 I FP16 Tensor	148 I 148 TFLOPS	119.5 I 119.5 TFLOPS	96.5 I 96.5 TFLOPS
TF32 Tensor	74 TFLOPS	59.8 TFLOPS	48.3 TFLOPS
FP32	44 TFLOPS	59.8 TFLOPS	24.1 TFLOPS
FP64	1 TFLOPS	N/A	N/A
RT Core	N/A	Yes	Yes
MIG	Up to 7 MIG	N/A	N/A
L2 Cache	60 MB	96 MB	36 MB
Media Engine	7 NVDEC, 7 NVJPEG	3 NVENC (+AV1), 3 NVDEC, 4 NVJPEG	2 NVENC (AVI), 4 NVDEC, 4 NVJPEG
Power	400 W	275W	TBD
Form Factor	8-way HGX	2-slot FHFL	1-slot LP
Interface	PCIe Gen5 x16: 128 GB/s	PCle Gen4 x16: 64 GB/s	PCle Gen4 x16: 64 GB/s
NVLink	900 GB/s	-	-
Samples	November 2023	November 2023	November 2023
Production	December 2023	December 2023	December 2023

Те са почти на границата на рестрикциите.

[emil vasilev]

NVIDIA създава аналог на RTX 4090, с който да избегне американските рестрикции за Китай. RTX 4090 има 660.8 * 8 = 5,286 TPP при FP8 и 330.4 * 16 = 5,286 TPP при FP16.FP е съкращение на Floating Point. Лимитът е 4800 TPP. Очаква се аналогът да бъде под този лимит.
NVIDIA Cooks Up The GeForce RTX 4090 D GPU, Brings Flagship Gaming Back To China With Full US Compliancy

[emil vasilev]

NVIDIA анонсира GeForce RTX™ 40 SUPER Series.
GeForce RTX 40 SUPER Series: New Heroes Debut in the Gaming and Creating Universe With AI as Their Superpower
NVIDIA Reveals Gaming, Creating, Generative AI, Robotics Innovations at CES
GeForce RTX 40 SUPER Series Graphics Cards Launching This January, For Supercharged Gaming & Creating, With Super-Fast AI

[emil vasilev]

NVIDIA пусна видеокарта за компактни работни станции --->NVIDIA RTX™ 2000 Ada Generation
NVIDIA RTX 2000 Ada Generation GPU Brings Performance, Versatility for Next Era of AI-Accelerated Design and Visualization
NVIDIA RTX™2000 Ada Generation

[emil vasilev]

NVIDIA пусна карти за мобилни работни станции RTX 500 и 1000 Ada Generation Laptop GPU.
Двете карти включват невронен процесорен модул (NPU), които осигуряват солидна мощност за AI приложения.
NVIDIA RTX 500 and 1000 Professional Ada Generation Laptop GPUs Drive AI-Enhanced Workflows From Anywhere
NVIDIA Professional Graphics Solutions