NVIDIA GeForce RTX 5090 and RTX 5080 Specs Leak: 21,760 Cores, 32GB VRAM, and Impressive Power Draw

RTX 5090 RTX 5090

NVIDIA’s upcoming GeForce RTX 5090 and RTX 5080 have been making headlines with leaked specifications that highlight their potential as formidable contenders in the GPU market. The anticipated flagship graphics cards are expected to offer substantial upgrades from their predecessors, creating excitement among PC hardware enthusiasts.

Short Summary:

  • The GeForce RTX 5090 slated to include an impressive 21,760 CUDA cores.
  • The RTX 5080 is projected to offer 10,752 CUDA cores with a lower power draw.
  • Both cards are expected to support cutting-edge technologies like PCIe 5.0 and DisplayPort 2.1a.

As we edge closer to NVIDIA’s anticipated release of the RTX 50-series, leaks have begun to surface regarding the specifications of the forthcoming GeForce RTX 5090 and RTX 5080 graphics cards. These revelations come from the reputable hardware leaker @kopite7kimi on X (formerly known as Twitter). According to the latest posts, these next-gen GPUs are shaping up to be significant upgrades over their predecessors — the RTX 4090 and RTX 4080.

images 2

The GeForce RTX 5090: A Powerhouse Awaits

Dominating the spotlight is the RTX 5090, which is reportedly built around the GB202-300-A1 Blackwell GPU. The specifications indicate a staggering 21,760 CUDA cores, representing a 33% increase over the RTX 4090’s 16,384 CUDA cores. Additionally, the new card is said to demand a powerful 600W total graphics power (TGP), a significant bump from the 450W required by the RTX 4090. This leap equates to a 33% rise in power consumption, ensuring that the RTX 5090 will deliver standout performance amidst the competitive landscape.

rtx 5090 leak 1

In terms of memory, the RTX 5090 is expected to house 32GB of GDDR7 VRAM on a 512-bit memory bus. This hardware configuration promises impressive specifications, with the potential for 1.7TB/s of memory bandwidth if the memory speeds reach an anticipated 28Gbps. Such capabilities would represent a striking 68% improvement compared to the previous flagship GPU.

The speculated impact on performance with these advancements cannot be overstated; users can expect substantial improvements in tasks that require high graphical fidelity. With this dramatically increased core count and upgraded memory bandwidth, gamers and content creators alike are poised to experience unparalleled performance in their applications.

The GeForce RTX 5080: Bridging the Gap

Alongside its flagship counterpart, the RTX 5080 has also garnered attention among hardware enthusiasts. Expected to employ a GB203-400-A1 GPU, this model is reportedly equipped with 10,752 CUDA cores, which surpasses the 9,728 CUDA cores found in the RTX 4080.

rtx 5080 leak 1

While the power requirements for the RTX 5080 are set at a reasonable 400W TGP, this marks an 80W increase over the RTX 4080’s 320W consumption. Prospective users will be excited to see that the RTX 5080 will also include 16GB of GDDR7 memory on a 256-bit memory bus, equivalent to what the current RTX 4080 Super offers, albeit with potential speed enhancements.

Technical Advancements: PCIe 5.0 and DisplayPort 2.1a

Both the RTX 5090 and RTX 5080 are slated to support cutting-edge technologies such as PCIe 5.0 and DisplayPort 2.1a connectivity. These advancements will ensure compatibility with the latest hardware and facilitate higher data transfer speeds. The PCIe 5.0 bus interface will likely enable enhanced performance efficiencies and help future-proof these cards for high-demand applications.

The introduction of DisplayPort 2.1a connectivity reveals NVIDIA’s intention to support high refresh rates and increased bandwidth capabilities, catering to the evolving needs of gamers and content creators alike.

Understanding Power Consumption and Connectors

With the RTX 5090’s increased power draw, both expected models will utilize a single 12V-2×6 PCIe power connector, alleviating concerns from dual connector setups. This change may simplify installation while maintaining necessary power levels. NVIDIA plans to fully utilize the specifications of the existing 16-pin power connection, which they first implemented with the RTX 4090.

Competitive Pricing Landscape

As we await further information about official pricing, speculation abounds regarding where the RTX 5090 and RTX 5080 will settle in the market. Historically, the RTX 4090 launched at a price of $1,599 while the RTX 4080 debuted at $1,199. Given these benchmarks, the RTX 5090 could command an even higher price tag due to its sheer performance potential.

Consequently, enthusiasts may need to factor in this cost when considering a new upgrade, as the performance per dollar could shift significantly with these new releases.

Industry Impact and Conclusion

The potential launch of the RTX 5090 and RTX 5080 later in 2025 is already generating buzz within the PC hardware community. Given the leaks, it’s clear NVIDIA aims to further solidify its dominance in the GPU market, particularly as competition from AMD is currently subdued. With outstanding specifications and new features, these cards aim to cater to a wide range of consumers, from gamers to professional content creators.

Staying tuned for official announcements and releases over the coming months will be essential for anyone involved in the PC building community. Consider utilizing resources such as comparison tools to find the best deals on compatible components when planning your upgrades.

As always, keep an eye on reliable sources as we build towards what could be one of the most exciting launches in GPU history. Stay connected with ThinkComputers.org for the latest updates and in-depth reviews.

Via kopite7kimi (X/Twitter)