#geforce_rtx_5090

[phonemantra-blog]

RTX 5090 will likely have fewer active cores The release of the GeForce RTX 5090 is still very far away, but gradually more and more information about this video card is appearing. Now, a very reliable insider Kopite7kimi has shared some characteristics of the GB202 GPU, which will form the basis of Nvidia’s new flagship. Judging by these data, the new GPU will receive 12 TPC blocks of eight SM blocks each. If the number of CUDA cores in one SM block does not change, and no rumors say anything about this, then the GB202 will receive 24,576 CUDA cores. For comparison, the […]

RTX 5090 will likely have fewer active cores The release of the GeForce RTX 5090 is still very far away, but gradually more and more information about this video card is appearing. Now, a very reliable insider Kopite7kimi has shared some characteristics of the GB202 GPU, which will form the basis of Nvidia’s new flagship. [caption id="attachment_61683" align="aligncenter" width="780"] GeForce RTX 5090[/caption] Judging by these data, the new GPU will receive 12 TPC blocks of eight SM blocks each. If the number of CUDA cores in one SM block does not change, and no rumors say anything about this, then the GB202 will receive 24,576 CUDA cores. For comparison, the AD102 has 18,432 CUDA cores. This will be the basis for the GeForce RTX 5090. At the same time, the RTX 5090 will likely receive GB202 in an incomplete configuration. For example, the RTX 4090 has 16,384 active CUDA cores, meaning 16 SM units are disabled, or about 11%. If the situation is approximately the same in the case of the new generation, then the RTX 5090 will receive approximately 170 SM blocks or 21,760 CUDA cores. Of course, these are all very rough estimates. It is also reported that the GB202 will receive a 512-bit memory bus. Whether the RTX 5090 or only the RTX 5090 Ti will receive it, if one comes out at all, is still anyone’s guess. But in any case, the GPU itself will have a much wider bus than the AD102. And if Nvidia uses GDDR7 memory in the new generation, the throughput of the RTX 5090 compared to the RTX 4090 will increase incredibly. Insider kopite7kimi was the first to accurately describe the parameters of GeForce RTX 30 cards, and he also described RTX 40 cards more than once. Overall, it is one of the most reliable sources in this segment.

[phonemantra-blog]

Even when delivering 640 W power Back in July, we learned that Nvidia was working on a replacement for the fire-hazardous 12VHPWR connector. And now the first tests of the 12V-2×6 connector have appeared on the Internet, which should replace the 12VHPWR. The authors of the HardwareBusters channel visited the laboratory of Linewell, which produces power cables, in particular for video card manufacturers. The GeForce RTX 5090 will not repeat the fate of the fire-hazardous RTX 4090. Linewell tested the cable with the new connector and tested it in a format that simulated the most common cause of 12VHPWR melting. At first, Linewell […]

Even when delivering 640 W power Back in July, we learned that Nvidia was working on a replacement for the fire-hazardous 12VHPWR connector. And now the first tests of the 12V-2x6 connector have appeared on the Internet, which should replace the 12VHPWR. [caption id="attachment_59926" align="aligncenter" width="780"] GeForce RTX 5090[/caption] The authors of the HardwareBusters channel visited the laboratory of Linewell, which produces power cables, in particular for video card manufacturers. The GeForce RTX 5090 will not repeat the fate of the fire-hazardous RTX 4090. https://youtu.be/aI-soO7_l_M Linewell tested the cable with the new connector and tested it in a format that simulated the most common cause of 12VHPWR melting. At first, Linewell specialists tested it in normal mode for almost an hour, and the connector only heated up to 46.5 degrees. Then the cable was not connected completely, which usually leads to melting, and a power of more than 640 W was applied to it, that is, exceeding the established limit of 600 W. The connector passed the test, heating up to only 41 degrees.

[phonemantra-blog]

The new product will receive 50% more CUDA cores The first data about GeForce RTX 50 video cards is beginning to appear on the Internet. In particular, the first rumors concern a model that can be called RTX 5090 in absentia, since it will be the successor to the current RTX 4090. GeForce RTX 5090 will be 70% faster than RTX 4090. So, insider Panzerlied, who was the first to share leaks about the parameters and performance of the RTX 4080 and RTX 4070 Ti, claims that from the RTX 5090, we can expect an increase in the number of […]

The new product will receive 50% more CUDA cores The first data about GeForce RTX 50 video cards is beginning to appear on the Internet. In particular, the first rumors concern a model that can be called RTX 5090 in absentia, since it will be the successor to the current RTX 4090.  [caption id="attachment_56753" align="aligncenter" width="780"] GeForce RTX 5090[/caption] GeForce RTX 5090 will be 70% faster than RTX 4090. So, insider Panzerlied, who was the first to share leaks about the parameters and performance of the RTX 4080 and RTX 4070 Ti, claims that from the RTX 5090, we can expect an increase in the number of computing units relative to the RTX 4090 by about 50%, and an increase in memory bandwidth by 52%, 78% increase in GPU cache (probably L2 cache), 15% increase in frequency. Collectively, this will lead to a performance increase of approximately 70%, which is very impressive compared to the already super-efficient RTX 4090.  Let us remember that long before the release of the RTX 4090, numerous rumors and leaks attributed it to approximately a twofold increase relative to its predecessor. As a result, in practice, the increase was about 50–70%, depending on which card is considered the direct predecessor of the RTX 4090 (RTX 3090 or RTX 3090 Ti). If the story is about the same as the RTX 5090, then we can expect a real increase of about 40–50%, which is also very, very significant.