NVIDIA GeForce RTX 5090 PCB Leak Exposes Massive GB202 “Blackwell” GPU Die, Samsung GDDR7 Memory Chips & Tons of VRMs

Hassan Mujtaba
NVIDIA GeForce RTX 5090 PCB Leak Exposes Massive GB202 "Blackwell" GPU Die, Samsung GDDR7 Memory Chips & Tons of VRMs 1

NVIDIA's GeForce RTX 5090 PCB has been leaked once again, this time fully equipped with the Blackwell GB202 GPU & GDDR7 memory.

NVIDIA's GeForce RTX 5090 Has One Crowded PCB That Features A Massive Blackwell GB202 GPU, Lots of GDDR7 Modules & A Power Delivery Designed For A Flagship

Yesterday, we got our first look at the PCB of the upcoming NVIDIA GeForce RTX 5090, which will be the flagship GPU within the RTX 50 "Blackwell" gaming lineup. The compact yet crowded PCB looks very powerful and would feature an insane number of components. Now, we have a new leak that once again gives us a better look at the PCB.

Related Story NVIDIA Blackwell Sweeps Every MLPerf 6.0 Benchmark With No Competition In Sight, While GB300 Systems Run Up to 60% Faster Than GB200

Spotted over at Baidu Forums, the new picture shows the PCB fully decked out with components and the three most important things to talk about are the GPU, the VRAM, and the VRMs. Starting with the GPU itself, the chip is labeled as GB202-300-A1, which confirms previous leaks, and we can also see that this is not a retail SKU but a qualification sample that should be close to the final design. The chip looks as massive as ever and has a very large IHS around it to ensure maximum contact with the heatsink.

PCB of the NVIDIA GeForce RTX 4090 GPU for comparison. (Image Source: NVIDIA)

Surrounding the GPU are 16 GDDR7 memory modules from Samsung which should be running at speeds of 28 Gbps. The RTX 5080 will be the only SKU in the lineup to feature higher GDDR7 DRAM speeds. These GDDR7 modules are arranged exactly as the past leaks had hinted and will feature speeds of up to 28 Gbps across a 512-bit bus interface. This would offer a total bandwidth of 1792 GB/s, a massive uplift over the 1008 GB/s bandwidth of the existing RTX 4090 GPU.

As for the VRMs, we can count a total of 28 VRMs within the section of the PCB pictured. There are a lot more capacitors based on the previous leak and the card will receive power through a single 16-pin (12V-2x6) connector interface. The NVIDIA GeForce RTX 5090 is expected to feature a TBP of up to 600W.

The NVIDIA GeForce RTX 5090 will be the flagship GPU and will launch either by the end of January or early February, while the RTX 5080 is currently scheduled to hit shelves by mid-January. Expect more information on the next-gen lineup at CES 2025, which is a few weeks away.

NVIDIA GeForce RTX 50 GPU Specs (Official):

Graphics Card NameNVIDIA GeForce RTX 5090NVIDIA GeForce RTX 5080NVIDIA GeForce RTX 5070 TiNVIDIA GeForce RTX 5070NVIDIA GeForce RTX 5060 TiNVIDIA GeForce RTX 5060NVIDIA GeForce RTX 5050
GPU NameBlackwell GB202-300Blackwell GB203-400Blackwell GB203-300Blackwell GB205-300Blackwell GB206-300Blackwell GB206-250Blackwell GB207-300
GPU SMs170847048363020
GPU Cores217601075289606144460838402560
Clock Speeds2017 MHz (Base)
2407 MHz (Boost)
2295 MHz (Base)
2617 MHz (Boost)
2295 MHz (Base)
2452 MHz (Boost)
2325 MHz (Base)
2512 MHz (Boost)
2407 MHz (Base)
2572 MHz (Boost)
2280 MHz (Base)
2497 MHz (Boost)
2317 MHz (Base)
2572 MHz (Boost)
Memory Capacity32 GB GDDR716 GB GDDR716 GB GDDR712 GB GDDR78-16 GB GDDR78 GB GDDR78 GB GDDR6
Memory Bus512-bit256-bit256-bit192-bit128-bit128-bit128-bit
Memory Speed28 Gbps30 Gbps28 Gbps28 Gbps28 Gbps28 Gbps20 Gbps
Bandwidth1792 GB/s960 GB/s896 GB/s672 GB/s448 GB/s448 GB/s320 GB/s
Power Interface1 12V-2x6 (16-Pin)1 12V-2x6 (16-Pin)1 12V-2x6 (16-Pin)1 12V-2x6 (16-Pin)16-Pin or 8-Pin8-Pin8-Pin
Launch30th Jan, 202530th Jan, 202520th Feb, 20254th March, 202516th April, 202519th April, 20251st July, 2025
TBP575W360W300W250W180W145W130W
Price (MSRP)$1999$999$749$549$379-$429$299$249
Which NVIDIA GeForce RTX 50 GPU are you looking forward to the most?

News Source: Harukaze5719

Hassan Mujtaba Photo

About the author: A Software Engineer by training and a PC enthusiast by passion, Hassan Mujtaba serves as Wccftech's Senior Editor for hardware section. With years of experience in the industry, he specializes in deep-dive technical analysis of next-generation CPU and GPU architectures, motherboards, and cooling solutions. His work involves not only breaking news on upcoming technologies but also extensive hands-on reviews and benchmarking.

Follow Wccftech on Google to get more of our news coverage in your feeds.

Button