TECH NEWS – Nvidia’s flagship card, based on the Blackwell architecture, will reportedly be capable of clock speeds that we haven’t seen much in graphics cards before.
On the Chiphell forum, Panzerlied, who previously posted accurate information about Nvidia’s GeForce RTX 5000 series cards, mentioned something again, but this time about the top-of-the-line model. Officially, “the greens” haven’t talked much about the new generation yet, but Jensen Huang, the company’s president, said during his Q&A at Computex that they will be “something” and he’s looking forward to telling us more about them. But what will the RTX 5090 be particularly muscular at?
Rumor has it that the graphics card’s clock speed will be close to 2.9GHz by default, so it won’t be far from 3GHz. It’s worth comparing it to the current flagship RTX 4090, which is clocked at 2235 MHz by default, so we’re looking at a 30% boost in that regard alone. We don’t know yet what the boost clock speed will be, but looking at the current generation, it will definitely be over 3 GHz (the RTX 4090’s is 2520 MHz on the reference card, but it could be 2.8 or even 2.9, maybe even 3 GHz by individual manufacturers with the right power consumption and cooling system).
The GeForce RTX 5090 will most likely use Blackwell’s GB202 GPU with a slightly modified chip, as it originally had a 512-bit memory bus and 192 SM (streaming multiprocessor), but this will be trimmed to 448 bits and 28 GB of GDDR7 VRAM. Of course, these are subject to change, and probably downward, because Nvidia hasn’t really brought out the real top chip in the last few generations (it usually ends in 0, and GB202 doesn’t…), so they’re holding back on what the technology could really be capable of.
The GeForce RTX 5090 and 5080 are supposed to come out towards the end of the year, but it’s also possible that Nvidia could push the announcement to early 2025. That would be unfortunate: Intel and AMD are about to launch their new processors!
Leave a Reply