SiS Precision and KYEC, which provide outsourced semiconductor assembly and test (OSAT) services for NVIDIA, said they have not been notified of NVIDIA’s increased order requirements. The news comes just days after one of NVIDIA’s newly announced RTX 40 series graphics cards, the GeForce RTX 4070, hit the market.

There are multiple theories to explain NVIDIA’s approach. The GeForce RTX 4070 is available at a suggested retail price starting at CNY 4799, and while it offers many new features and performance improvements, it not only has to compete with AMD’s current graphics cards such as the RX 6800, which has recently dropped in price to $469 but also with the previous generation GeForce RTX 30 series graphics cards that are still in stock. RTX 30 series graphics cards are still in stock. Some believe that NVIDIA’s strategy is to maintain its current production demand in order to clear out its previous generation inventory. Meanwhile, the newly released RTX 4070, like the previously released RTX 4080, is overwhelmingly well-supplied with non-public versions.

Another theory is that NVIDIA is more focused on its enterprise graphics cards, such as the A100, which are being purchased in large quantities by multiple entities for AI training and chatbots, causing NVIDIA’s stock to continue to soar.

Regardless of the reason, NVIDIA still expects to release more RTX 40 series graphics cards in the coming months.

The post NVIDIA is not in a hurry to increase the production of the GeForce RTX 40 series GPU appeared first on TechGoing.

According to reports, the performance of RTX 4050 is 7% stronger than RTX 3060; RTX 4060 is 24% stronger than RTX 3060, slightly weaker than RTX 3070; RTX 4070 is 37% stronger than RTX 3060, and 5% stronger than RTX 3070 Ti.

In terms of parameters, the RTX 4070 laptop GPU has 4608 CUDA cores with a frequency of up to 2175 MHz and is equipped with 128-bit wide 8GB GDDR6 video memory. The RTX 4060 GPU has 3072 CUDA cores, the GPU frequency can reach 2370MHz, and it is equipped with 8GB 128bit GDDR6 video memory.

For comparison, the old RTX 3070 GPU has 5120 CUDA cores, equipped with 8GB 256bit video memory; the old RTX 3060 GPU has 3840 CUDA cores, equipped with 6GB 192bit video memory.

The post NVIDIA RTX 40 notebook GPU performance: 4050 is slightly stronger than 3060 appeared first on TechGoing.

Now, Kopite, who has a good record of breaking the news, says that the RTX 40 series does not have a Titan model.

Nvidia had launched a TITAN RTX model in the RTX 20 era with 24GB of video memory, while the contemporary RTX 2080 Ti had only 11GB of video memory. In the RTX 30 era, the RTX 3090 had 24GB of video memory, and Nvidia did not launch a TITAN model. Therefore, in the RTX 40 era or even later, Nvidia is unlikely to launch TITAN, and if you want more memory, you can only choose professional cards.

According to reports, Nvidia’s best AD102 chip has been reserved for the RTX 4090 Ti, which will have 18176 CUDA cores, which is 11% more than the 16384 CUDA of the RTX 4090. At the same time, the RTX 4090 will be equipped with faster 24 Gbps 24GB of GDDR6X memory, and the bandwidth will increase from 1 TB/s to 1.1 TB/s. Taken together, the RTX 4090 Ti will perform 10% to 20% higher than the RTX 4090.

The post Nvidia RTX 40 series does not have a 48GB memory Titan model appeared first on TechGoing.

According to Nvidia, the current RTX 4080 graphics card is priced from $899 (about RMB 6400.88), while the RTX 3060 is priced from $329 (about RMB 2342.48). Thus, it seems that the RTX 3060-80 series cards will continue to exist for some time, together with the RTX 4090 and 4080 to form the latest ecology.

The post NVIDIA’s latest RTX graphics card family: RTX 4090/4080 + RTX 3080/3070/3060 appeared first on TechGoing.

The NVIDIA RTX 30 series already supports AV1 hard decoding but does not support AV1 hard coding, and this feature is finally completed on the RTX 40 series.

In 2018, the Open Media Alliance (AOMedia) released a new generation of video codec AV1 (AOMedia Video Codec 1.0). Developed in collaboration with the Open Media Alliance, the encoding encodes 4K UHD video with an average 30% higher compression rate than comparable encoders. Previously, only Intel Ruixuan supported AV1 codec, and neither Nvidia nor AMD fully supported it.

Nvidia said in a press release that the GeForce RTX 40 series uses the eighth-generation Nvidia Video Encoder, or NVENC for short, and now supports AV1. The new GeForce RTX 4090 brings a huge performance boost, 3rd Gen RT Cores, 4th Gen Tensor Cores, 8th Gen NVIDIA Dual AV1 encoder and 24GB of Micron G6X memory capable of 1TB/s bandwidth. GeForce RTX 4090 is 2x faster than GeForce RTX 3090 Ti in 3D rendering, AI and video export.

Nvidia RTX 4090 will be available on October 12, and RTX 4080 will be available in November.

The post NVIDIA RTX 40 series graphics cards finally support AV1 hardware codec appeared first on TechGoing.

As shown in the picture above, with the blessing of RTX 4090, “Rainbow Six Siege” can reach 1440p 493fps, and “Fortnite” can reach 1440p 601fps.

NVIDIA said that NVIDIA has brought great changes to the competitive gaming industry in recent years: GeForce RTX graphics cards have increased frame rates to levels previously unimaginable; NVIDIA and panel partner AU Optronics have worked together to create and promote high refresh rate, no tearing G-SYNC gaming monitors; in addition, NVIDIA Reflex provides gamers and developers with a full suite of innovative tools that profoundly changes the way system latency is detected and reduced. Currently, almost all professional esports players and events use esports monitors with GeForce RTX graphics cards with refresh rates above 240Hz.

With the GeForce RTX 40 series refreshing the concept of “extreme speed” once again, NVIDIA says, e-sports players can upgrade their gear to a 27-inch 1440p 360Hz display, enjoy stable 360 ​​FPS picture quality, and experience as low as NVIDIA Reflex games. 10ms system latency.

Nvidia research found that players can aim up to 3% better on a 1440p 27-inch monitor than a traditional 1080p 24-inch monitor when aiming at small targets such as an enemy’s head. For competitive players and professional e-sports players, 3% will be the key to winning or losing.

The NVIDIA RTX 4090 will be available on October 12, and the RTX 4080 will be available in November.

The post NVIDIA RTX 40 series graphics cards announced, 1440p 360 FPS gaming era is coming appeared first on TechGoing.

This is almost comparable to the GH100 core for accelerated computing, the latter is also TSMC 4nm process, integrated up to 80 billion transistors, AD102 is only a mere 5 billion short.

In addition, RTX 30 series large core GA102 integrated 28.3 billion transistors, AD102 is as much as 2.65 times it, is the largest increase in recent years.

Further ahead, 12nm TU102 18.6 billion, 12nm GV100 21.1 billion, 16nm GP102 12 billion, their densities are 25 million per square millimeter, GA102 significantly increased to 45.1 million per square millimeter.

AD102 core integrated 18,432 CUDA cores, compared to GA102 increased by about 71%, the memory specifications are basically the same, I do not know where so many transistors are used. Is there a secret weapon?

Now the suspense depends on how large the area of the AD102 core, how high the degree of integration.

The post NVIDIA’s 4nm RTX40 graphics card will have 75 billion transistors appeared first on TechGoing.

It is said that the RTX 40 graphics card using TSMC 5nm process GPU frequency compared to the RTX 30 series using Samsung 8nm process has been greatly improved. RTX 4070 base frequency of 2310MHz, Boost frequency up to 2610MHz, the maximum frequency of ultra 2800MHz.

For comparison, Nvidia’s current RTX 3070 has a Boost frequency of 1.73GHz.

The NVIDIA RTX 40 series graphics cards are expected to be released in September or October.

The post TSMC 5nm to power, the news that Nvidia RTX 4070 graphics card Boost frequency 2.6GHz appeared first on TechGoing.