Nvidia 5000 Series Launch – Release Date, Specs and Performance

---

Are you on the verge of buying your next Nvidia GPU this month? Then this one’s for you. There are strong rumors circulating in the GPU news platforms that the release of the next-gen Nvidia 5000 series is just around the corner. This news was originally reported by Quasarzone, a source known for its strong Korean connections and considered one of the most reliable in the industry.

In this blog post, we’ll discuss in detail what gamers, content creators, and everyday PC users should expect from the RTX 5000 series, and is it worth waiting for the new launch. So, keep reading till the end if you don’t want to opt for a GPU that you might regret buying later. Let’s dive in!

Nvidia RTX 5000 Release Date

According to Korean resources, Nvidia has lowered the production of its current generation GPUs. They have informed their partners that they will drastically lower their shipments of RTX 40 cards. This proved that they are cooking the Nvidia 5000 series launch. Furthermore, the rumors suggest that you should expect the possible Nvidia RTX 5000 release date to be in the second half of the current year.

Reasons for Quick Nvidia 5000 Series Launch

One of the key reasons for this early launch is that the Super variant (super cards) didn’t perform well in the market. Gamers are not super-excited to buy the mid-range CPU. Moreover, GeForce RTX 4070 Super was a “disaster” as quoted by Nvidia insiders.

As a result, GeForce RTX 5090 release date is on the cards and is all set to shake the market with a bang.

However, the question remains the same: What should we expect from this upcoming launch?

You may also like: The Top Graphics Cards for Gamers in 2023: A Comparative Review

Nvidia 5000 Series – Rumored Features and Performance Upgrades

As for now, we don’t have any concrete information; however, there are a few rumors regarding the performance upgrades and improvements.

Less Power Consumption with MCM Design Approach 

Nvidia may stick to its single-chip approach for graphics cards, as it always does. However, there’s a twist: they’re considering a new method called Multi-Chip Module (MCM). This approach involves using separate chips connected to one substrate. It’s like aligning all the blocks together to come up with something bigger and much better performance-wise.

According to some insiders, the next data center chips, GB100 and GB102, might use this MCM design. Moreover, regular users utilizing high-end GB202 might also leverage this feature.

However, Nvidia plans to stick with the usual single-chip setup for cheaper variants like GB203, GB205, GB206, and GB207 chips.

GDDR7 Graphics Memory – A Heaven for Gamers

One of the major reasons that the gamers and content creators have not gone crazy buying the Ada Lovelace lineup is the same GDDR6 and GDDR6X technology utilization. But, now there are strong rumors that the Nvidia 5000 series GPU will upgrade to the latest GDDR7 technology, which are great signs.

To make it more clear, take the example of GeForce RTX 4090 and older versions of RTX 3090 Ti! They both had a speed of 1,008 GB/s. On the other hand, the anticipated memory speed for the Nvidia 5090 variant is 1,536 GB/s, which will be a massive jump of 52%!

• Memory Configuration

There’s been a lot of discussion about the memory bus of the GeForce 5090. Initially, it had a 512-bit memory configuration. But later, Nvidia changed the configuration to match the 384-bit bus of the Nvidia 4000 and 3000 series GPUs. However, the company reverted it to a 512-bit bus for the top-tier Blackwell GPU in March 2024.

• VRAM Sizes

Additionally, rumors suggest that the GB205 and GB203 won’t have the same configuration as the flagship model. They might have 128-bit and 256-bit buses respectively. Regarding VRAM size, there’s speculation that the RTX 5090 may have 36GB of GDDR7 memory, but it is still a rumor.

Still, we expect it to have at least 24GB, similar to the RTX 4090. There’s also talk that the second-tier GB203 card might have the same memory size as the GB202 variant, potentially meaning an RTX 5080 could have anywhere from 12 to 18GB of memory.

You may also like: Nvidia RTX A4000: A Deep Dive into Its Features and Benefits

Higher Count of CUDA Cores (3nm Manufacturing Process)

Nvidia keeps upgrading their GPUs, and the Blackwell series is no different. They’re using a newer, smaller manufacturing process this time, called TSMC’s 3nm. This is smaller than the Ada Lovelace series’ 5nm process.

According to a tweet by kopite7kimi, we may not see a big jump in the number of Texture Processor Clusters (TPCs) and Graphics Processing Clusters (GPCs) on the new GPUs. Another leak suggests that apart from a 78% in cache and a 15% boost in frequency, the RTX 5090 might have a 50% improvement in the overall scale. This may result in a higher CUDA core count, somewhere around 24,576 cores.

DisplayPort 2.1 and PCIe 5.0 Compatibility

The upcoming GeForce RTX 5000 graphics cards are said to improve display capabilities compared to older models. Moreover, kopite7kimi said that these new GPUs would work with the DisplayPort 2.1 standard, resulting in the following:

• Higher resolution
• Faster refresh rate
• Better bandwidth

While Ada Lovelace GPUs can be used with PCIe 5.0 motherboards, they can’t maximize the super-fast connection speed. However, Blackwell GPUs are expected to fully utilize the lightning-quick data transfer rates of PCIe Gen 5.0.

Wrapping Up

Whether the launch happens this year or in 2025, one thing is certain: the Nvidia 5000 series will probably be a great hit. We may see an insane response from the buyers. Stay tuned to Direct Macro for further product details and updates if you want to buy the best processors at the best prices.