
Zotac Accidentally Confirms Nvidia RTX 50 Series Ahead of CES 2024 Launch | Image Source: www.tomshardware.com
TAIPEI, Taiwan, Dec. 16, 2024 — In a surprising yet revealing development, Zotac has inadvertently unveiled key details about Nvidia’s highly anticipated RTX 50 series GPUs weeks ahead of their official announcement. According to Tom’s Hardware, Zotac mistakenly listed entries for the RTX 5090, RTX 5080, and RTX 5070 family of GPUs on its website. The listings, which were quickly removed, confirm critical specifications, including Nvidia’s flagship RTX 5090 featuring 32GB of GDDR7 VRAM. These details shed light on Nvidia’s preparations for its next-generation graphics cards, likely set to debut at CES 2024.
Zotac’s misstep, reported initially by Videocardz, marks the third instance of the company prematurely listing GPUs before launch. However, this accidental leak provides the first quasi-official confirmation of several key features and product lineups for Nvidia’s upcoming RTX 50 series. The incident also reinforces growing speculation about the performance improvements and hardware advancements Nvidia plans to introduce with its next-gen Blackwell architecture.
RTX 5090: Confirmed to Feature 32GB of GDDR7 VRAM
The inadvertent listings by Zotac reveal that the RTX 5090, Nvidia’s next flagship GPU, will include 32GB of GDDR7 memory, a significant upgrade from the 24GB GDDR6X memory featured on the current RTX 4090. According to Tom’s Hardware, Zotac’s updated filtering options for its graphics card catalog explicitly included parameters like “GDDR7” and “32GB,” aligning with earlier leaks about the RTX 5090’s enhanced memory capacity.
As per reports, the RTX 5090 is expected to be powered by Nvidia’s GB202 chip, a massive GPU die reportedly over 744mm² in size, packing up to 170 Streaming Multiprocessors (SMs) out of a possible 192. For comparison, the RTX 4090 featured 128 SMs. The massive 32GB VRAM configuration and architectural improvements suggest that Nvidia’s upcoming GPU will target the most demanding workloads, such as 8K gaming, AI processing, and advanced content creation.
Interestingly, Zotac’s listings also included a China-specific variant, the RTX 5090D. While details remain scarce, the RTX 5090D is expected to feature the same GB202 GPU but with fewer enabled SMs—reportedly around 150. This would likely make the RTX 5090D a slightly nerfed version of the flagship model, tailored to regional market demands.
RTX 5080 and RTX 5070 Family: Details Revealed
Further down the product stack, Zotac’s listings highlighted the RTX 5080 and RTX 5070 series. According to Tom’s Hardware, the RTX 5080 is expected to utilize Nvidia’s GB203 silicon, featuring 84 SMs, which is less than half the SM count of the RTX 5090. The GPU will reportedly include 16GB of GDDR7 memory, positioning it as a strong contender in the high-end gaming market but at a lower price point than the flagship model.
The RTX 5070 family, which includes both the RTX 5070 and RTX 5070 Ti, is also expected to use the GB203 chip. As stated by Kopite, a well-known hardware leaker, the RTX 5070 Ti is likely to feature 70 SMs. However, details about the base RTX 5070 remain sparse, with rumors suggesting a release timeline for February 2024. It is unusual for Nvidia to launch Ti variants at the beginning of a GPU generation, as these typically arrive as mid-cycle refreshes. Nevertheless, the inclusion of the Ti version in Zotac’s listings suggests Nvidia might take a different approach this time around.
TSMC 4NP Process and Blackwell Architecture
The RTX 50 series GPUs are widely expected to leverage TSMC’s advanced 4NP process node, a refined version of its 5nm-grade technology. According to Tom’s Hardware, this new manufacturing process offers 30% higher transistor density compared to the 4N process used in Nvidia’s current RTX 40 series. The improvements in transistor density are likely to translate into better power efficiency and higher performance per watt for the RTX 50 lineup.
Nvidia’s Blackwell architecture, which will power the RTX 50 series, is also rumored to introduce native support for new data types such as FP4 and FP6, similar to Nvidia’s server-grade GPUs like the B100 and B200. While these changes are primarily aimed at AI and data center workloads, their inclusion in consumer GPUs could open new possibilities for AI-powered applications, ray tracing performance, and upscaling technologies like DLSS.
While Nvidia has yet to confirm the specifics, the architectural changes and increased memory capacities suggest that the RTX 50 series will offer significant performance gains over its predecessors. These improvements could make the RTX 50 lineup particularly appealing to gamers, content creators, and AI developers looking for cutting-edge graphics processing capabilities.
What to Expect at CES 2024
The timing of Zotac’s accidental listings aligns with Nvidia’s expected unveiling of the RTX 50 series at CES 2024 in January. Jensen Huang, Nvidia’s CEO, is widely anticipated to take the stage and introduce the new GPUs, detailing their specifications, pricing, and release timelines. As per industry insiders, the RTX 5090 will likely lead the lineup, with the RTX 5080 and RTX 5070 family following closely behind.
CES has historically been a major event for Nvidia, serving as the launchpad for its most significant GPU announcements. Given the scale of improvements expected with the RTX 50 series, the event could mark a turning point in the GPU market, particularly as competition from AMD and Intel continues to heat up. Nvidia’s focus on GDDR7 memory, advanced manufacturing processes, and architectural upgrades underscores its commitment to maintaining its leadership in the high-performance GPU segment.
In line with the trends seen in previous generations, pricing and availability will be critical factors for Nvidia’s RTX 50 series success. While the flagship RTX 5090 is expected to command a premium price, the RTX 5080 and RTX 5070 series could provide more affordable options for gamers and enthusiasts. The early leaks have already generated significant excitement, and all eyes will be on Nvidia’s announcements at CES next month.
In conclusion, Zotac’s inadvertent leak has confirmed several key details about Nvidia’s RTX 50 series GPUs, including the flagship RTX 5090 with 32GB of GDDR7 memory and the RTX 5080 and RTX 5070 families. With Nvidia poised to unveil its next-generation GPUs at CES 2024, the RTX 50 series promises substantial performance upgrades driven by advanced manufacturing processes, improved architecture, and higher memory capacities. As the industry awaits Jensen Huang’s keynote, the leaked specifications hint at a new era of GPU performance and innovation.