Home Tech NVIDIA’s groundbreaking next-generation GeForce GPU graphics card rumored to feature AD102, 21Gbps, 24GB, up to 600W TGP

NVIDIA’s groundbreaking next-generation GeForce GPU graphics card rumored to feature AD102, 21Gbps, 24GB, up to 600W TGP

0
NVIDIA’s groundbreaking next-generation GeForce GPU graphics card rumored to feature AD102, 21Gbps, 24GB, up to 600W TGP

NVIDIA’s Ada Lovelace GPUs will be available to gamers later this year in next-generation GeForce RTX graphics cards. These cards are expected to rock some extreme specs and some extreme GPU requirements to work as detailed before Kopite7kimi In his latest round of leaks.

Detailed NVIDIA Next Generation GeForce Graphics Card: AD102 GPU, 24GB Memory, 21Gbps, up to 600W TGP

There may be a file new rumor Surrounding the naming convention used by next-generation GeForce graphics cards, some are now referring to the name change to the GeForce RTX 50 series instead of the previously expected GeForce RTX 40 series. In the end, whatever NVIDIA decides to call the next GeForce lineup isn’t nearly as important as the base specs.

EVGA GeForce RTX 3090 Ti KINGPIN Hybrid Graphics Card rocks dual 16-pin connectors for over 1200W

NVIDIA GeForce RTX 4090 ‘Ada Lovelace’ graphics card with ‘rumored’ GA102-300 GPU specification

We know from leaked NVIDIA Ada Lovelace GPU configurations that the flagship GA102 Gaming chip will rock a total of 144 SMs, up 71% over the current GA102 GPU, and featuring 18,432 CUDA core counts. The chip used by the main GeForce graphics card will rock the AD102-300 GPU and we can’t say for sure if this will be a full-fat SKU since the RTX 3090 Ti with the GA102-350 rocks the full configuration while the GA102 -300 SKU featured in the NVIDIA GeForce RTX 3090 is a shortened SKU code.

According to Kopite7kimi, the flagship GeForce graphics card will be based on the PG137/139-SKU30 PCB. These are the reference designs and the founders version. The reference design is provided to AIBs as a reference for their custom models while the FE’s design is exclusive to the Founders Edition reference model.

The graphics card is said to maintain a memory capacity of 24 GB while providing a memory speed of 21 Gbps via the 384-bit bus interface. This is the same memory design as the flagship Ampere GeForce, GeForce RTX 3090 Ti, and should rock the same amount of bandwidth at 1004 Gbit/s. As for the power input, the card will have a single PCIe Gen 5 connector with a TGP rating of up to 600W. You’ll need a 4×8-pin to 1x-16-pin adapter if you’re running an older PSU or a single Gen 5 to Gen 5 connector if you have a compatible power supply.

NVIDIA Ada Lovelace ‘GeForce RTX 40’ GPU Configurations

GPU name GPCs / TPCs SMS per TPC/Total koda cores L2 cache Memory bus
M 102 12/6 2/144 18432 96 MB 384 bit
M 103 7/6 2/84 10752 64 MB 256 bit
M 104 5/6 2/60 7680 48 MB 192 bit
M 106 3/6 2/36 4608 32 MB 128 bit
M 107 3/4 2/24 3072 32 MB 128 bit

NVIDIA Ada Lovelace & Ampere GPU Comparison

Ada Lovelace GPU short message koda cores TOP SKU Memory bus GPU amp short message koda cores TOP SKU Memory bus SM increase (% over ampere)
M 102 144 18432 RTX 4090? 384 bit GA102 84 10752 RTX 3090 Ti 384 bit + 71%
M 103 84 10752 RTX 4070? 256 bit GA103S 60 7680 RTX 3080 Ti 256 bit + 40%
M 104 60 7680 RTX 4060? 192 bit GA104 48 6144 RTX 3070 Ti 256 bit + 25%
M 106 36 4608 RTX 4050 Ti? 128 bit GA106 30 3840 RTX 3060 192 bit + 20%
M 107 24 3072 RTX 4050? 128 bit GA107 20 2560 RTX 3050 128 bit + 20%

The NVIDIA Ada Lovelace GPU family is expected to take a generation leap similar to Maxwell’s for Pascal. It is expected to launch in the second half of 2022 but we expect the offering and pricing to be similar to the current cards though NVIDIA Spending Billions of Dollars to Get Good TSMC 5nm Chips.

NVIDIA CUDA GPU (Rumoured) Introductory:

GPU TU102 GA102 M 102
building Turing Ampere Ada Lovelace
to treat TSMC 12nm NFF Samsung 8nm TSMC 5 nm
die size 754 mm 2 628 mm 2 ~ 600 mm 2
Graphics Processing Kits (GPC) 6 7 12
Textile Processing Kits (TPC) 36 42 72
multiprocessor flow (SM) 72 84 144
koda cores 4608 10752 18432
L2 cache 6 MB 6 MB 96 MB
Theoretical TFLOPs 16.1 37.6 ~90 TFLOPs?
memory type GDDR6 GDDR6X GDDR6X
Memory bus 384 bit 384 bit 384 bit
Memory capacity 11 GB (2080T) 24 GB (3090) 24 GB (4090?)
Major SKU RTX 2080 Ti RTX 3090 RTX 4090?
TGP 250 watts 350 watts 450-850 watts?
Release September 2018 September 20 2H 2022 (TBC)

What next generation GPUs are you most looking forward to?

LEAVE A REPLY

Please enter your comment!
Please enter your name here