TECH NEWS – While it’s common to see graphics cards that take up at least two slots in your computer (and even three or four slots are standard now), Nvidia has gone back to the past.

6144 CUDA cores, 20 GB of GDDR6 VRAM, 130W TDP – that’s how you can describe the RTX 4000 graphics card. No, we’re not talking about the RTX 4000 family of GeForce cards (for gaming), and not the RTX 4000 SFF Ada Generation graphics card (which is powerful as hell among low profile cards!), but simply called RTX 4000, because Nvidia’s naming scheme is incomprehensible. In the past, cards for workstations were branded Quadro, but the “green” company dropped the name under the Turing architecture (RTX 2000 series)…

The card with the Ada Lovelace AD104 chip was not designed by Nvidia for gaming (but can be used for that) and is between the RTX 4070 and 4070 Ti in performance. However, while it has much more VRAM, its bandwidth is smaller. 160-bit results in 360 GB/s, with fewer cores than the RTX 4000 GPU in workstation laptops. Therefore it is pretty similar to the RTX 4000 SFF Ada Generation, although that consumes only 70W (!!!).

The RTX 4000 Ada Generation (officially its full name) is unique because it occupies a single card slot, so it doesn’t hang onto the other PCI-E slots with its black and gold design and the blower style fan (which therefore lets hot air out the back). Via the 12-pin connector, it consumes 130W and has 26.7 TFLOPs (the RTX 4070 reaches 29.15, despite having fewer cores and VRAM). Another factor that might make it a terrible gaming choice is the price: it will be available from September for $1250, so it will only be found in corporate machines. (However, it could be a great buy used in a few years!)

The $2250 RTX 4500 is coming in October, and the RTX 5000 is available now for $4000…

Source: PCGamer