Currently, Nvidia’s flagship game graphics card is RTX 4090, and its memory interface is 384bit.

According to WccfTech reports, since Nvidia released the GeForce GTX 285 graphics card in 2009, Nvidia’s GeForce gaming graphics card series has not provided a memory configuration above 384 bits. The RTX 4090 and RTX 6000 Ada with 384bit bit width use 24 GB and 48 GB of video memory. In the future, Nvidia’s new generation of flagship game cards and workstation graphics cards can easily achieve 32GB and 64GB video memory with the help of 512bit bit width. If the latest 24Gb (3GB) GDDR7 video memory particles are used, the video memory capacity can even reach 48GB and 96GB.

Samsung has announced that its GDDR7 video memory can reach a speed of 32 Gbps, which can achieve a bandwidth of 2 TB / s with a 512bit bit width.

It was earlier reported that Nvidia’s Ada-Next flagship gaming GPU is expected to be available sometime in 2025.

The post Nvidia’s next-gen Ada-next flagship graphics card supports 512bit bit wide video memory appeared first on TechGoing.

It is noticed that Nvidia changes its architecture roughly every 24-48 months. The current RTX 40 series graphics cards use the “Ada Lovelace” architecture. Renewed through 2025.

▲ Picture source hardwareLUXX

According to the newly released roadmap, the next-generation accelerator card structure “Hopper Next” will become the mainstream of Nvidia in the first half of 2024. It was previously reported that the chip, code-named “Blackwell”, will be used in Nvidia’s next-generation accelerator cards in the future. The current H100 accelerator card based on the Hopper architecture has been well-received by users. Therefore, the successor “Hopper Next” may be hotly anticipated,

The next-generation game architecture “Ada Lovelace Next” is scheduled to debut in 2025 and is expected to be used in the RTX 50 series. According to foreign media videocard, Nvidia is expected to release more mid-term facelift “Ti” models of the RTX 40 series by 2025.

▲ Picture source videocardz

In addition, since Nvidia has now fully focused on the new architecture, foreign media hardwareLUXX said that there will be no new generation of GeForce products in the second half of 2023 and 2024.

Nvidia predicts that Grace and Grace Hopper will become the mainstream of the industry in the second half of 2023, and Nvidia is currently shipping its Grace and Grace Hopper professional CPUs in large quantities. The next-generation “Grace Next” is scheduled to be released in 2025. Nvidia’s new roadmap around today is expected to push its HPC and AI supercomputers forward.

The post Nvidia is expected to launch RTX 50 Series graphic cards in 2025 appeared first on TechGoing.

(Source: Nvidia Technical Blog)

It is reported that the Ada Lovelace GPU’s Optical Flow Accelerator (NVOFA) component can apply the same Optical Flow logic to video applications and graphics rendering.

While predicting the next frame, the GPU can also generate the frame through AI to improve the smoothness of the picture – what Nvidia calls “engine-assisted frame rate conversion” (FRUC).

We’ve seen other forms of “fluidity-enhancing” solutions on some TVs (and AMD graphics drivers) before, but here NVENC also compares two real frames in a video, determines motion vectors, and sets up an optical flow stage to ensure the accuracy of the generated frames.

Finally, Nvidia will release FRUC as a library to help it better integrate into content creation and media consumption applications.

The post NVIDIA Brings New Video Framerate AI Boosting Experience to Ada Lovelace GPUs appeared first on TechGoing.

The RTX 40 series Ada Lovelace GPU has been replaced with TSMC’s new 4N process node. Even if the default 2.6 GHz main frequency of the public version card, its energy efficiency and frequency improvements are quite gratifying.

As for AIB partners, we expect more SKUs that are factory overclocked at 2.7 – 2.8 GHz, but are actually expected to go above 3 GHz.

For reference, the non-public GPU of the RTX 30 series Ampere RTX 3090 Ti is clocked at around 1.8 – 1.9 GHz, but it can actually exceed 2 GHz in games.

It should be pointed out that such an exaggerated performance increase also means that the RTX 40 series Ada Lovelace GPU needs to consume a lot of power and generate more heat at the same time.

However, the Founders Edition public PCB + heatsink design shown by Green Factory left a deep impression on us as always.

Nvidia said that it further optimized the air duct of the dual axial fan, increased the fan size and fan blade volume by 10%, and upgraded the RTX 4090 to the 23-phase power supply.

In addition, thanks to lower memory temperatures, the newer, more powerful Ada Lovelace GPUs are able to run cooler with the help of stock coolers and retain the excellent overclocking potential for gamers.

Finally, the RTX 40 series FE graphics cards all use PCIe 5.0 16-pin connectors that meet the new generation ATX 3.0 power specification to reduce cable clutter caused by the old triple 8-pin connectors.

In the case of using high-quality power cables, players can only use a single connector to meet the power consumption requirements of RTX 40 series graphics cards. In addition, many commercial power supply manufacturers have also opened application links for 16-pin adapter cables.

It is reported that brands including ASUS, Cooler Master, FSP, MSI, ThermalTake, iBuyPower, etc., will launch a new generation of ATX 3.0 power supplies in October this year.

While the Nvidia AIB graphics partner has not shared the final frequencies and TGP/TBP power consumption of their non-public designs, we can expect to hear more in the coming days.

The post Nvidia claims Ada Lovelace GPUs like RTX 4090 can be overclocked above 3GHz appeared first on TechGoing.

Hassan Mujtaba retweeted the tweet and explained that Bernoulli’s number is the iconic algorithm used in analytic engines and was proposed by Ada Lovelace in 1842.

While we’ve been talking about the Ada Lovelace GPU architecture for a long time, it’s great that Nvidia continues to draw inspiration from the historically renowned scientists whose inventions have laid an important foundation for the current technology world.

In many ways, Ada Lovelace can be considered the world’s first computer enthusiast – she was the first to realize that the analytical engine proposed by Charles Babbage had applications beyond pure computing A computer programmer’s first set of algorithms.

Such an advanced idea is nearly half a century earlier than the work done by the “father of computer science” Alan Turing during the World Wars and the invention of the general-purpose computer.

In fact, as early as during the GTC 2018 keynote, NVIDIA has listed the comic avatars of several famous physicists, mathematicians, and scientists.

If all goes well, we will be welcoming various GeForce RTX 40 series graphics cards with Ada Lovelace GPU cores in the coming months, with the RTX 4090 flagship SKU expected to arrive first on October 22nd.

The post NVIDIA Confirms GeForce RTX 40 Series GPUs Will Use Ada Lovelace Naming appeared first on TechGoing.