The GeForce RTX 4080 graphics card from NVIDIA has received a new update to its specifications from leaker Kopite7kimi, who claims even faster GDDR6X memory. The leaker claims that all other parameters for the NVIDIA GeForce RTX 4080 graphics card are essentially unchanged, except memory and power design.
It appears that NVIDIA will make use of Micron’s faster GDDR6X memory modules; the 4080 is said to have 23 Gbps dies. The TBP increased somewhat as a result of this specification modification, going from 320W to 340W.
In comparison to the previous version, which offered 80 SMs or 10,240 cores, the NVIDIA GeForce RTX 4080 is anticipated to use a scaled-down AD103-300 GPU configuration with 9,728 cores or 76 SMs enabled of the total 84 units.
Due to its condensed design, the RTX 4080 might only have 48 MB of L2 cache and fewer ROPs than the complete GPU, which has 64 MB of L2 cache and up to 224 ROPs. The PG136/139-SKU360 PCB is anticipated to serve as the card’s foundation.
In terms of memory specifications, the GeForce RTX 4080 is predicted to use 16 GB of GDDR6X memory, clocked at 23 Gbps via a 256-bit bus interface. This will offer bandwidth of up to 736 GB/s. Given that it has a 320-bit interface but only a meager 10 GB capacity, this is still a little bit slower than the 760 GB/s bandwidth provided by the RTX 3080. NVIDIA might be adding a next-generation memory compression suite to make up for the 256-bit interface to make up for the reduced bandwidth.
The TBP’s power rating has been increased by 20W to 340W, up from the 320W specification we previously received. This puts the TBP on par with the current RTX 3080 graphics card (up to 350W). Although it is unknown if the other RTX 40 series graphics cards will also benefit from the faster GDDR6X memory upgrade, Micron has already started full-scale production of up to 24 Gbps GDDR6X memory modules, so they must be used someplace.
Also read: Intel Meteor Lake CPUs will remortedly feature Ray Tracing Support
