NVIDIA GeForce RTX 5090 PCB Leak Reveals Massive GB202 GPU, 16 GDDR7 Memory Modules & Single 16-Pin Power Connector

Recently, a leak from the Chiphell Forums unveiled the PCB design of NVIDIA’s upcoming GeForce RTX 5090 graphics card. This marks a significant moment for enthusiasts eagerly awaiting the new generation of NVIDIA GPUs centered around the RTX 50 series.

Key Features of the NVIDIA GeForce RTX 5090 PCB

The NVIDIA GeForce RTX 5090 is poised to be the flagship model in the RTX 50 “Blackwell”GPU range. While specifications have been speculated and discussed in the recent past, this latest leak provides a detailed view of the PCB, indicating it may represent a reference design from AIB partners.

At the forefront of this leak is the substantial BGA package for the cutting-edge GB202 GPU, which will serve as the cornerstone for the RTX 5090 and potential variants like the RTX 5090 Ti or RTX 5080 Ti. According to the information disclosed, the RTX 5090 will utilize the GB202-300 GPU architecture featuring an impressive configuration of up to 21,760 CUDA cores spread across 170 streaming multiprocessors. Excitingly, expected clock speeds could approach 3 GHz under standard conditions.

Memory Architecture and Bandwidth

The memory capabilities of the RTX 5090 are equally noteworthy, equipped with 16 GDDR7 memory slots that allow for a maximum of 32 GB of VRAM. This aligns with previous speculations and will utilize GDDR7 modules that reach speeds of up to 28 Gbps via a 512-bit bus interface. This configuration potentially elevates total memory bandwidth to a staggering 1792 GB/s, markedly surpassing the 1008 GB/s offered by the RTX 4090. Additionally, the PCB is speculated to include around 30 VRM phases and over 60 capacitors, providing robust power management.

NVIDIA GeForce RTX 5090 Render — Image Source: Wccftech

Powering the RTX 5090 appears straightforward, with a single 16-pin power connector adhering to the latest 12V-2×6 standard. This design allows the GPU to draw up to 600W, aligning with the expected total board power (TBP) for this model.

NVIDIA GeForce RTX 5090 PCB Details — Image Source: Chiphell Forums

A closer look at the PCB reveals the model number “PNT5090 v1.0″along with “100-10145-DFAA-A02,”confirming that NVIDIA will provide two distinct baseboard designs: the PG144 for the Founders Edition and PG145 for AIB models. Notably, this PCB is more compact than earlier models, suggesting a return to a more standardized design philosophy compared to previous generations.

Anticipated Launch of the RTX 50 Series

The NVIDIA GeForce RTX 5090 is expected to launch between late January and early February 2025, while the RTX 5080 is slated to arrive in mid-January. Keep an eye on the upcoming CES 2025, where further revelations about NVIDIA’s next-generation graphics lineup are anticipated.

Preliminary Specs for NVIDIA GeForce RTX 50 GPUs

Graphics Card Model	NVIDIA GeForce RTX 5090	NVIDIA GeForce RTX 5080	NVIDIA GeForce RTX 5070 Ti	NVIDIA GeForce RTX 5070	NVIDIA GeForce RTX 5060 Ti	NVIDIA GeForce RTX 5060
GPU Name	Blackwell GB202-300	Blackwell GB203-400	Blackwell GB203-300-A1	Blackwell GB205-300-A1	Blackwell GB206?	Blackwell GB206?
Streaming Multiprocessors (SMs)	170 (192 Full)	84 (84 Full)	70 (84 Full)	50 (50 SM Full)	TBD	TBD
CUDA Cores	21,760 (+33%)	10,752 (+11%)	8,960 (+16%)	6,144	TBD	TBD
Clock Speeds	TBD	TBD	TBD	TBD	TBD	TBD
L2 Cache	TBD	TBD	TBD	TBD	TBD	TBD
Memory Capacity	32 GB GDDR7 (+33%)	16 GB GDDR7 (0%)	16 GB GDDR7	12 GB GDDR7	16 GB GDDR7	8 GB GDDR7
Memory Bus Width	512-bit (+33%)	256-bit (0%)	256-bit (0%)	192-bit (0%)	128-bit (0%)	128-bit (0%)
Memory Speed	28 Gbps	30 Gbps	28 Gbps	28 Gbps	28 Gbps?	28 Gbps?
Bandwidth	1792 GB/s	1024 GB/s	896 GB/s	672 GB/s	448 GB/s	448 GB/s
Total Board Power (TBP)	600W (+33%)	400W (+25%)	300W (+10.5%)	250W (+14%)	TBD	TBD
Power Interface	1 x 12V-2×6 (16-Pin)	1 x 12V-2×6 (16-Pin)	1 x 12V-2×6 (16-Pin)	1 x 12VHPWR (16-Pin)	1 x 12VHPWR (16-Pin)	1 x 12VHPWR (16-Pin)
Expected Launch	Q1 2025	Q1 2025	Q1 2025	Q1 2025	TBD	TBD