The strategy with 5080 looks quite similar. It obtains a slight increase in the nucleus count (10.5 percent more than the original 4080 and only 5 percent over 4080 Super), plus the same 16 GB of RAM in the same 256 -bit memory interface. However, changing from GDDR6x to GDDR7 gives you a decent increase with memory bandwidth, however, in the order of just over 30 percent compared to the Super 4080 and 4080 Super.
As with the 5090, the 12 -pin feed connector is now slightly recessed and at an angle so that it does not protrude so far in its case.
Credit: Andrew Cunningham
But all that is incredibly soft and incremental compared to the big jumps through the dashboard in the nucleus count, the width of the memory bus, the size of the memory and the bandwidth of the memory you get when you pass from the RTX 4090 to RTX 5090. Basic counts and memory bandwidth are also maintained well below RTX 4090. This has the benefit of not exploiting the power requirements of 5080 (at 360 W, it is only 40 W higher than 4080, and in our real evidence, the 5080 actually did not consume much more power on average, but also means that 4090 and 5090 continue to separate from the rest of the alignment, inaccessible to anyone with less than a couple of thousands of dollars to spend on a GPU.
As for the physical design of the version of the Edition of Founders of Nvidia of 5080, it is identical to that of 5090, just as 4080 resembled 4090. 5080 feels lighter than 5090, which makes sense, a Card with much more under energy consumption and half of the CUDA nuclei do not need the same type of cooling apparatus. But the dimensions and connectors on both cards are exactly the same.
GPU temperatures measured during reference runs.
Credit: Andrew Cunningham
In our tests, 5080 worked a little warmer than 4080 under load, but only for approximately 5 ° Celsius. The 5080 Fundadores edition works with 11 ° or 12 ° colder than the edition of the 5090 Founders, so it should be a bit easier to maintain its fresh system with one of these installed.