Nvidia and AMD both are gearing up for next-Gen of graphics cards. While the last generation had a lot of potential, graphics cards were plagued by supply issues coupled with increased demand due to cryptocurrency mining and extreme price hikes. The situation has gradually improved, and the industry has now set its eyes on the upcoming graphics cards from both Nvidia and AMD camps. Going by the current reports, both brands will be launching products with significant performance uplifts over the last generation.

Nvidia’s Next-generation Ada Lovelace RTX 4000 series GPUs are rumored to be extremely power-hungry. We have already covered leaks regarding Nvidia’s upcoming GeForce RTX 4080. The GeForce RTX 4080 is rumored to be based on TSMC’s 5nm process node. Furthermore, AD102 GPU Core (RTX 4090 and up) is also rumored to consume up to 600 watts TBP (Total Board Power). On the other hand, AMD’s RDNA 3 lineup looks much more efficient in power draws. AMD’s flagship, Navi 31 is said to consume significantly less power for similar performance levels.

A recent tweet from trusted leaker @kopite7kimi claims that a full-fat AD102 SKU exists with a 900W TGP. 

However, Greymon55 has recently stated that the 900 watts is probably an overstatement. Gerymon55 is also highly reliable and well known in the industry. 

As the tweet states, a 900-watt GPU will most likely not be a product from Nvidia. However, Greymon has clarified that the GeForce RTX 4090 Ti will consume twice as much power as the Navi31 (Radeon RX 7900 XT series). Nvidia’s power draws have been a matter of concern as of late. Sources have claimed that this extreme power draw results from Nvidia’s attempts to push the new GPUs to their limits. 

Greymon also referenced an older tweet when talking about the recent developments. The tweet seems to hint at Nvidia’s greater heat output along with the extremely high power ratings compared to AMD’s RDNA3.

AMD’s RNDA 3 line-up Alleged Specifications: 

GPU Name Navi 21 Navi 33 Navi 32 Navi 31
GPU Process 7nm 6nm 5nm/6nm 5nm/6nm
GPU Package Monolithic Monolithic MCM MCM
Shader Engines 4 2 4 (2 per GCD) 6 (3 per GCD)
GPU WGPs 40 20 40 (20 per GCD) 60 (30 per GCD)
SPs Per WGP 128 256 256 256
Compute Units (Per Die) 80 40 80
160 (Total)
120
240 (Total)
Cores (Per Die) 5120 5120 5120 7689
Cores (Total) 5120 5120 10240 15360
Memory Bus 256-bit 128-bit 192-bit 256-bit
Memory Type GDDR6 GDDR6 GDDR6 GDDR6
Infinity Cache 128 MB 128-256 MB 384 MB 512 MB
Flagship SKU Radeon RX 6900 XTX Radeon RX 7700 XT? Radeon RX 7800 XT? Radeon RX 7900 XT?
TBP 330W ~200W ~300W ~400W
Launch Q4 2020 Q4 2022? Q4 2022? Q4 2022?

 

GeForce RTX 4000 line-up Alleged Specifications: 

Ada Lovelace GPU SMs CUDA Cores Top SKU Memory Bus Ampere GPU SMs CUDA Cores Top SKU Memory Bus SM Increase (% Over Ampere)
AD102 144 18432 RTX 4090? 384-bit GA102 84 10752 RTX 3090 Ti 384-bit +71%
AD103 84 10752 RTX 4070? 256-bit GA103S 60 7680 RTX 3080 Ti 256-bit +40%
AD104 60 7680 RTX 4060? 192-bit GA104 48 6144 RTX 3070 Ti 256-bit +25%
AD106 36 4608 RTX 4050 Ti? 128-bit GA106 30 3840 RTX 3060 192-bit +20%
AD107 24 3072 RTX 4050? 128-bit GA107 20 2560 RTX 3050 128-bit +20%

 

It seems like Nvidia has decided to go with a brute force approach with the upcoming GPUs. We already saw them trying to push the GA102 die to its limits with the GeForce RTX 3090Ti. Perhaps that was just Nvidia’s way of testing the waters in terms of power limits. Nonetheless, AMD seems to be keeping up in terms of performance while outclassing Nvidia regarding efficiency. As the date for the GPU reveals draws nearer, more leaks are expected to come out and we, are looking forward to learning more about the next generation.

More News: Apple Arcade’s PixelJunk Scrappers Could Be Coming To PC 

News reporter with years of experience who is creative and motivated. Excellent narrative abilities and a keen sense of the truth. Highly enthusiastic about the gaming and tech industry with a passion for writing.