There’s no other words for it, NVIDIA’s newest entry-level graphics card is a severely underpowered lemon and is a direct insult of the intelligence of their customer base.
Previously, we discussed NVIDIA’s decision to discontinue production of the RTX 3050 8GB model. This move was prompted by concerns that its sales were impacting the market share of the RTX 4060, despite both cards being similarly priced but offering vastly different performance levels as one would expect.
Now, mere months later, NVIDIA has unveiled their latest entry-level offering: the RTX 3050 6GB. This component features 10% less cores and a significantly reduced 96-bit memory interface versus the 8GB model under the same name scheme, utilizing previous generation hardware, all for the price of just $180.
https://www.nvidia.com/en-au/geforce/news/rtx-3050-6gb-out-now/
The motive behind their initiative becomes clear: the RTX 3050 8GB was priced at $230, as the RTX 4000 series seemingly begins at an outlandish $300 price point.
However, despite a substantial 62% performance gap between the RTX 3050 and RTX 4060, NVIDIA chose not to offer affordable hardware but instead continued producing outdated components for lower-value segments, aiming to extract maximum profits from the consumer.
The revamped RTX 3050 6GB continues this trend, possibly repurposing unsold mobile devices featuring the downsized GA107 core. Priced at a more accessible $180, the RTX 3050 6GB is geared towards offering a significantly inferior gaming experience for supposedly greater profits value.
Contrary to its name, the RTX 3050 6GB has undergone significant cuts in hardware specifications beyond just its memory capacity. Clock speed, power draw, memory bus width, and memory capacity have all been significantly reduced, making this card more akin to an RTX 3040 rather than a variant of the RTX 3050 8GB.
TechPowerUp noted that the RTX 3050 6GB stands out as the fastest GPU on the market that does not require a dedicated supplementary power connector, but that is a far cry from the reality surrounding this product.
The lack of a need for an external PCIe power connector is the sole redeeming quality of this product. This feature enables effortless drop-in upgrades for bothersome OEM pre-built PCs and compact low-power builds, making it a convenient option for certain users.
However, defending such a feature is utterly incomprehensible. NVIDIA could have effortlessly introduced a current-generation RTX 4050 as a mainstream offering with a TDP lower than 75 watts.
Despite the RTX 4050 Mobile’s meager 96-bit memory interface and 6GB of GDDR6 VRAM, it would undeniably offer consumers a low-priced, low-power option that outperforms the previous generation’s RTX 3050 6GB with little to no effort.
Moreover, it would provide support for NVIDIA’s artificially locked DLSS 3 frame generation, elevating the overall user experience.
NVIDIA seems content with delivering the bare minimum while maximizing revenue. The exorbitant prices of the RTX 4000 series, coupled with the subpar hardware they offer versus previous generations, serve as clear evidence that consumers are being taken advantage of.
In comparison to the RTX 3050 8GB, the RTX 3050 6GB features 11% fewer CUDA cores and SMs, a 21% lower boost clock, 33% less memory bandwidth and capacity, but boasts an 85% lower TDP. However, its actual performance doesn’t just fall short of expectations.
It’s an outright flop.
According to TechPowerUp’s benchmarks, the RTX 3050 8GB outperforms the RTX 3050 6GB by an average of 28% at 1080P, 30% at 1440P, and 40% at 4K resolution, though those latter two resolutions are completely irrelevant for this market segment.
Furthermore, the RTX 3050 6GB falls behind NVIDIA’s GTX 1660 Ti 6GB in all three resolutions, which is fucking staggering given how the GTX 1660 Ti released five years ago on February 22nd 2019.
According to TechPowerUp’s testing, the five-year-old Turing-based GTX 1660 Ti exhibits a remarkable 18% higher average performance at 1080P, 20% at 1440P, and 15% at 4K resolution.
As mentioned earlier, the RTX 3050 6GB is priced at $180 USD. In comparison, Intel’s ARC Alchemist A580, which features 8GB of GDDR6 VRAM across a 256-bit memory interface, is priced at only $165.
Surprisingly, the ARC Alchemist A580 outperforms all of NVIDIA’s offerings, boasting an average of 54% higher performance at 1080p compared to the RTX 3050 6GB.
While I’m not a math expert, a 54% boost in overall gaming performance for $15 less seems like a significantly better deal to me. Despite Intel’s immature driver support issues, I would unquestionably choose the A580 over NVIDIA’s dogshit offering any day of the week.
On TechPowerUp’s list, the only cards that the Nvidia RTX 3050 6GB could surpass were the older NVIDIA GTX 1060 6GB and AMD’s RX 580 8GB, with both the Pascal and Polaris architectures releasing back in 2016.
Interestingly, there are two specific titles where the RTX 3050 6GB is outperformed by the aging AMD Radeon RX 580: Marvel Spider-Man Remastered and Capcom’s Resident Evil 4 Remake.
This highlights the underwhelming performance of the RTX 3050 6GB, as it can be surpassed by products in the ~$200 price range that were released eight years ago.
The RTX 3050 6GB does also outperform the Radeon RX 6500 XT 4GB. It’s worth noting that the GTX 1650/1650 Super and GTX 1660/1660 Super were not part of this comparison. However, judging from TechPowerUp’s results, it’s probable that the RTX 3050 6GB would offer similar performance to the GTX 1650 Super given how far off the mark it is versus the 1660 Ti.
When it comes to low-end entry-level hardware, consumers often defend NVIDIA’s anti-consumer practices and products that offer significantly inferior performance and value compared to competitors across various price brackets, simply due to the fact that these products provide “better software” and “features” such as NVIDIA’s proprietary DLSS upscaling technology that noticeably impedes raw image clarity in favor of performance and of course Ray Tracing capabilities.
When considering Ray Tracing, it’s hard to overlook the fact that the only graphics card on the market capable of handling such a demanding task at a somewhat acceptable level of performance is the RTX 4090.
However, NVIDIA’s primary marketing strategy with the GeForce RTX rebranding undeniably revolves around ray tracing, so therefore entry-level GPUs should apply right?
Due to the even greater memory limitations, the RTX 3050 8GB is shown to be 36% faster at 1080p resolution, the RTX 3060 is shown to provide nearly double the performance, and for shits and giggles NVIDIA’s flagship GeForce RTX 4090 is nearly seven times faster with a 596% increase.
Needless to say, enabling ray tracing with the RTX 3050 6GB is out of the question. Alan Wake 2, for instance, struggles to achieve performance in the single-digit range at 1080p. Similarly, titles like A Plague Tale Requiem, Cyberpunk 2077, Elden Ring, F1 23, Ratchet & Clank, and Hogwarts Legacy typically deliver anywhere from 15 to 28 frames per second on average which are entirely unplayable.
The only area where the RTX 3050 6GB showed competitiveness was in TechPowerUp’s efficiency results. According to the review, the 3050 6GB demonstrated efficiency levels comparable to Nvidia’s Ada Lovelace GPU architecture and was only 3% less efficient than the RTX 4060.
Furthermore, the Ampere GPU exhibited 32% higher efficiency compared to the RTX 3050 8GB. This outcome is logical, considering the lower clock speeds and fewer GDDR6 memory modules of the 6GB variant, likely enabling the GPU to operate within a more efficient voltage curve.
However, TechPowerUp’s review confirms that the RTX 3050 6GB’s performance is absolute dogshit, making it arguably the worst graphics card available in the entry-level sub $200 price point.
According to TechPowerUp’s benchmarks, AMD’s RX 6600, priced just $20-$30 higher, offers a whopping 60% improvement in performance. The same holds true for Intel’s ARC A750, and as previously mentioned you can save $15 and get a 54% performance boost with Intel’s ARC A580.
The sole saving grace of the RTX 3050 6GB lies in its 75W power envelope, making it compatible with systems that lack a dedicated supplementary PCIe power cable.
Regrettably, in such setups, the 3050 6GB is often the only feasible choice, as NVIDIA appears hesitant to offer consumers a suitable low-power variant like the RTX 4050 and your only other alternatives would be the Radeon RX 6400 which isn’t even worth mentioning and a Ryzen 8000G APU which does not apply as that would be an entire system upgrade.
The reluctance of a desktop variant of the RTX 4050 is likely driven by NVIDIA’s preference for the higher profits generated by mobile devices equipped with the GeForce RTX 4050.
Purchasing this product is a clear sign of mental malnourishment as its niche market is evident. Despite being the top choice among GPUs without an external power connector, the glaring truth remains.
NVIDIA could effortlessly provide a vastly superior product at the same price point. Hence, the RTX 3050 6GB, more fittingly dubbed as the RTX 3040, stands as a woefully inadequate and pitiful waste of sand and plastic, reminiscent of AMD’s disappointing Radeon RX 7600.
