NVIDIA GeForce RTX 4070 SUPER vs NVIDIA RTX 3000 Ada Generation Laptop GPU vs NVIDIA GeForce RTX 4060 Ti 16G

NVIDIA GeForce RTX 4070 SUPER

The Nvidia GeForce RTX 4070 SUPER is a high-end desktop graphics card based on the Ada Lovelace architecture. It uses the AD104 chip and offers 7,168 cores and 12 GB GDDR6X graphics memory (192-bit bus with 21 Gbits).

The performance of the RTX 4070S lies between the older RTX 4070 and RTX 4070 Ti in the top class of high-end GPUs in 2024. Compared to AMD GPUs, the 4070 Super is similar fast as the AMD Radeon RX 7800 XT. This makes the GPU best suited for QHD and maximum details of demanding games (such as Avatar Frontiers of Pandora, Allan Wake 2, Cyberpunk 2033 or F1 23). Less demanding games can also be played smoothly in 4K (AC Mirage, Cod MW 3, ...).

The AD104 chip is produced at TSMC in 5 nm (4N with EUV) and is specified with a TGP (Total Graphics Power) of 200 watts.

NVIDIA RTX 3000 Ada Generation Laptop GPU

The Nvidia RTX 3000 Ada Generation, not to be confused with the A3000, P3000 or RTX 3000 Turing Generation, is a higher-end professional graphics card for use in laptops that sports 4,608 CUDA cores and 8 GB of ECC GDDR6 VRAM (error correction can be turned off if desired). It would be fair to say that this is a GeForce RTX 4070 Laptop in disguise; consequently, the graphics card is fast enough for many games at 1440p with quality set to High. Brought into existence in 2023, the RTX 3000 leverages TSMC's 5 nm process and Nvidia's Ada Lovelace architecture to achieve very decent performance combined with moderate power consumption. The Nvidia-recommended TGP range for the card is very wide at 35 W to 140 W leading to bizarre performance differences between different systems powered by what is supposed to be the same product.

Quadro series graphics cards ship with much different BIOS and drivers than GeForce cards and are targeted at professional use rather than gaming. Commercial product design, large-scale calculations, simulation, data mining, 24 x 7 operation, certified drivers - if any of this sounds familiar, then a Quadro card will make you happy.

Architecture and Features

Ada Lovelace brings a range of improvements over older graphics cards utilizing the outgoing Ampere architecture. It's not just a better manufacturing process and a higher number of CUDA cores that we have here (up to 16,384 versus 10,752); under-the-hood refinements are plentiful, including an immensely larger L2 cache, an optimized ray tracing routine (a smarter technique is used to determine what is transparent and what isn't), and other changes. Naturally, these graphics cards can both encode and decode some of the most widely used video codecs, AVC, HEVC and AV1 included; they also support a host of Nvidia technologies, including Optimus and DLSS 3, and they can certainly be used for various AI tasks.

The RTX 3000 Ada features 36 RT cores of the 3rd generation, 144 Tensor cores of the 4th generation and 4,608 CUDA cores, making it ~11% inferior to the RTX 3500 Ada Generation as long as we disregard clock speed differences. (The 3500 also has more VRAM). The RTX 3000 comes with 8 GB of 128-bit wide ECC GDDR6 memory for throughput of ~256 GB/s. The fact that error correction is present here proves that the graphics adapter is indeed targeted at professional users, and it can be disabled if desired.

Just like Ampere-based cards, the RTX 3000 makes use of the PCI-Express 4 protocol. 8K SUHD monitors are supported, however, DP 1.4a video outputs can potentially prove to be a bottleneck down the line.

Performance

Having reviewed the Dell Precision 5480 and its 50 W RTX 3000 Ada in September 2023, we can firmly say the graphics card is in the same league as the RTX 4060 Laptop and the RTX 4050 Laptop, despite sharing most of its specifications with the RTX 4070 Laptop. 48 seconds in the Blender 3.3 Classroom CUDA test and a 3DMark 11 Performance GPU score of less than 30,000 points are nothing to write home about. The GPU's gaming talents are not mind-blowing, either, although it is certainly capable enough for 1080p gaming.

• Tiny Tina's Wonderlands - almost 90 fps at 1080p / High
• Cyberpunk 2077 - over 60 fps at 1080p / High

Nvidia's marketing materials mention "up to 20 TFLOPS" of performance, a very noticeable improvement over 14.5 TFLOPS delivered by the RTX 2000 Ada Generation. Your mileage may vary depending on how competent the cooling solution of your laptop is and how high the TGP power target of the graphics card is. A 100 W RTX 3000 Ada might be able to come close to the average RTX 4070 Laptop.

One other thing worth mentioning is that enabling error correction appears to reduce the amount of video memory that is available to applications and games by up to a gigabyte.

Power consumption

Nvidia no longer divides its laptop graphics cards into Max-Q and non-max-Q models. Instead, laptop makers are free to set the TGP according to their needs, and the range can sometimes be shockingly wide. This is especially the case for the RTX 3000, as the lowest value recommended for it sits at just 35 W while the highest is 300% higher at 140 W (this most likely includes Dynamic Boost). This is the kind of delta that we've be seeing on consumer-grade laptops featuring the latest GeForce RTX cards.

Last but not the least, the improved 5 nm process (TSMC 4N) the RTX 3000 Ada is built with makes for very decent energy efficiency, as of mid 2023.

NVIDIA GeForce RTX 4060 Ti 16G

The Nvidia GeForce RTX 4060 Ti 16 GB is a mid-range desktop GPU and is manufactured using the 5 nm process at TSMC. In addition to the 8 GB model, there is now also a 16 GB model, which is primarily intended to offer noticeable added value for performance-hungry games. With the two GeForce RTX 4060 models, Nvidia is rounding off its current Ada-Lovelace line-up towards the mid-range. The chip used here is the AD106. This combines 4,352 memory units and 16 GB GDDR6 are available here. The memory interface is 128 bits, resulting in a memory bandwidth of 288 GB/s.

Features

The Nvidia GeForce RTX 4060 Ti 16 GB is not deprived of the features already introduced with the high-end models of the Ada-Lovelace generation, such as DLSS 3.0 with frame generation. This clearly sets the new mid-range model apart from the Nvidia GeForce RTX 3060 Ti and can gain a clear advantage in games that offer these features.

Performance

Without DLSS, the performance is comparable to that of an Nvidia GeForce RTX 4060 Ti 8 GB. Compared to an Nvidia GeForce RTX 3060 Ti, the Nvidia GeForce RTX 4060 Ti 16 GB can deliver around 10 percent more performance. Compared to AMD, the Nvidia GeForce RTX 4060 Ti 16 GB is slightly ahead of the Radeon RX 7600. The 16 GB VRAM is noticeable in some games. Sometimes we see significantly better frame rates in games where 8 GB VRAM is not enough. Nvidia has succeeded in its goal of creating a good gaming GPU for 1080p gaming and with 16 GB of RAM, the Nvidia GeForce RTX 4060 Ti is the more interesting of the two RTX 4060 Ti models. Unfortunately, Nvidia charges a hefty premium for the larger VRAM. This does not necessarily make it easier for the Nvidia GeForce RTX 4060 Ti 16 GB to position itself on the market.

Power consumption

The power consumption of the Nvidia GeForce RTX 4060 Ti 16 GB is specified by the manufacturer as 165 watts (TDP). This is 5 watts more than the 8 GB model. In order not to have to supply the graphics chip with less energy, an increase in power consumption was unavoidable. Compared to the Nvidia GeForce RTX 3060 Ti, however, the Nvidia GeForce RTX 4060 Ti 16 GB is still significantly more economical.