At CES 2025, NVIDIA made one of the most significant announcements, unveiling its latest 50-series RTX graphics cards. These new GPUs were accompanied by assertive promises, such as the RTX 5070’s performance that could rival the power of the RTX 4090 at a significantly lower cost.
Indeed, those assertions were significantly influenced by enhanced AI capabilities through DLSS 4, and NVIDIA failed to provide substantial performance data. This lack of solid numbers raised concerns among PC enthusiasts who are cautious about overemphasizing super-resolution and frame generation. In fact, I was also planning on upgrading to an RTX 5070 before the CES controversy, but in a different article discussing next-generation Radeon cards, I suggested that AMD could be a better choice for my personal PC setup.
At CES, NVIDIA’s Editor’s Day provided additional insights about the native RTX 5000 series performance versus earlier cards. Our trusted partners at Tom’s Hardware were present to gather information, offering a clearer picture of what we might anticipate from upcoming NVIDIA GPUs such as the RTX 5090, RTX 5080, RTX 5070 Ti, and RTX 5070.
NVIDIA shares native RTX 5000 performance comparisons
It appears that many people’s suspicions were confirmed, as NVIDIA’s performance claims during its CES 2025 keynote speech heavily emphasized the new Multi Frame Generation feature of DLSS 4, a technology exclusive to the latest Blackwell cards and their advanced architecture, boasting enhanced neural rendering capabilities. This has sparked curiosity about the actual native performance of these cards when DLSS is not activated.
In the presented slide, NVIDIA’s new graphics card, the RTX 5070, which I’m keen on and is expected to be highly sought-after at its $549 price point, is compared against its predecessor, the RTX 4070. This comparison was made by NVIDIA under a 1440p resolution with maximum settings.
When DLSS isn’t activated, or put simply, it uses only its native hardware capabilities, the RTX 5070 is expected to perform around 20% better than the RTX 4070 in gaming scenarios. This performance remains consistent when you add DLSS 3 without ray tracing. However, once DLSS 4 is activated, we witness significant enhancements in frame rates, even with full ray tracing turned on.
Given that the RTX 4070 debuted at a Manufacturer’s Suggested Retail Price (MSRP) of $599, and if the RTX 5070 delivers a 20% performance boost while reducing its price by 8%, it’s clearly beneficial for the RTX 5070.
The RTX 5070’s non-Ti counterpart has a comparable enhancement in performance compared to its previous RTX 4070 Ti. It’s noteworthy that when DLSS 4 is activated, performance soars even higher than the standard version, reaching a maximum boost of 2.9 times in Indiana Jones and the Great Circle.
As a tech enthusiast, I’m excited to share that the long-awaited RTX 4070 Ti hit the shelves at an MSRP of $799. Now, here’s some fantastic news – NVIDIA has decided to sweeten the deal even further! They’ve dropped the price for their upcoming RTX 5070 Ti by a cool $50! That means even more power and performance at an increasingly affordable cost. Can’t wait to see what else they have in store for us!
Shifting our focus towards the high-performance models, NVIDIA opted for a 4K resolution for the comparison of its RTX 5080 and RTX 5090.
Initially comparing the RTX 5080 with the RTX 4080, NVIDIA suggests a performance boost of approximately 15% when DLSS (Deep Learning Super Sampling) is turned off. The improvement remains significant when DLSS 4 is activated, with certain games showing a doubled performance ratio. Notably, both GPUs are slated to hit the same $999 MSRP as the RTX 4080. This pricing suggests that the jump from the RTX 5070-class cards may not be as substantial as initially perceived.
The RTX 5090 takes the lead in terms of comparisons; NVIDIA estimates a 30% performance boost over the RTX 4090 when DLSS is turned off, which is the highest leap among all the GPUs in the 50-series lineup. However, this top-tier performance doesn’t come at a low price tag.
As an analyst, I’ve been keeping a close eye on the upcoming release of the RTX 5090. While its predecessor, the RTX 4090, hit the shelves at $1,599, the RTX 5090 is set to retail at $1,999 – that’s a 22% price increase. However, this price hike comes with significant advancements. The RTX 5090 boasts enhanced Tensor Cores, brand-new Streaming Processors, and improved RT Cores compared to the RTX 4090. It also features 32GB of faster GDDR7 VRAM, a significant upgrade from the 24GB of GDDR6 in the RTX 4090.
NVIDIA seems to have latency in check
In simpler terms, Multi Frame Generation refers to the increased production of visual frames by artificial intelligence. However, this raises concerns about potential delays in delivering these frames. Regardless of how swiftly your Graphics Processing Unit (GPU) generates the frames, it won’t matter if they can’t be delivered promptly.
On a different slide, NVIDIA addressed concerns regarding latency. By using the RTX 5090 as an illustration, the company demonstrated a substantial decrease in latency when DLSS 4 is activated. This reduction went from approximately 135ms to about 30ms in Cyberpunk 2077, with Black Myth: Wukong not lagging too far behind in terms of improvement.
Everything that NVIDIA has disclosed here is truly valued, undeniably easing some concerns. Nevertheless, it’s crucial to be patient and look forward to actual user reviews to gauge the performance of the new RTX 5000 GPUs under real-world conditions.
Read More
- TNSR PREDICTION. TNSR cryptocurrency
- TRB PREDICTION. TRB cryptocurrency
- RLC PREDICTION. RLC cryptocurrency
- POL PREDICTION. POL cryptocurrency
- OKB PREDICTION. OKB cryptocurrency
- CTXC PREDICTION. CTXC cryptocurrency
- GBP EUR PREDICTION
- SEI PREDICTION. SEI cryptocurrency
- HOOK PREDICTION. HOOK cryptocurrency
- SXP PREDICTION. SXP cryptocurrency
2025-01-16 18:39