New details of NVIDIA GeForce RTX 4090 & RTX 4080 graphics cards have been revealed regarding PCB and memory capacity.

NVIDIA GeForce RTX 4080 Reportedly Featuring Two PCB Designs With 16 & 12 GB Capacity, RTX 4090 Gets 14-Layer PCB

According to information by MegasizeGPU, the NVIDIA GeForce RTX 4090 will come in one PCB flavor while the RTX 4080 will come in two flavors. The RTX 4090 will use the same PCB for reference and the AIC design will be equipped with up to 24 GB of GDDR6X memory and a total of 14 layers for the PCB. The NVIDIA GeForce RTX 4080 has leaked which you can read more about here.

The NVIDIA GeForce RTX 4080, on the other hand, will reportedly have two completely different PCB designs. The former is said to have up to 16 GB of GDDR6X memory capacity and a 12 layer design for AIC cards while the second model will feature 12 GB GDDR6X memory support & a 10 layer PCB design for AIC and reference models.

Now we have heard reports that 16 GB memory is being the official model and nothing about the 12 GB variant but NVIDIA does have two RTX 3080 models, one with 10 GB and the other with 12 GB, launched earlier. The 12GB RTX 3080 features a 384-bit bus while the RTX 3080 features a 320-bit bus.

Since the RTX 4080 is rumored to feature a 256-bit bus for the 16 GB variant, the 12 GB AIC model will feature a cut-down 192-bit bus interface which is a significant reduction and very low bus width for flagship graphics cards. . NVIDIA had to compensate for GPU bandwidth by increasing the memory clock and introducing new memory compression technology to make up for the lost bandwidth. Maybe that’s why the RTX 4080 was recently rumored to have received a 23 Gbps GDDR6X memory bump.

It’s no surprise to us that the 16/12 GB model is nothing more than an engineering sample as we’ve seen 16GB and 20GB RTX 3080 models in the wild. This graphics card is indeed made and manufactured by NVIDIA but is not fit for retail launch.

NVIDIA GeForce RTX 4080 Specs ‘Expected’

NVIDIA GeForce RTX 4080 It is expected to use the cut-down AD103-300 GPU configuration with 9,728 cores or 76 SM enabled out of a total of 84 units while the previous configuration offers 80 SM or 10,240 cores. While the full GPU comes with 64 MB of L2 cache and up to 224 ROP, the RTX 4080 might end up with a lower 48 MB of L2 cache and ROP too due to its cropped design. This card is expected to be based on the PG136/139-SKU360 PCB.

As for the memory specs, the GeForce RTX 4080 is expected to have a capacity of 16 GB GDDR6X which is said to be adjustable at 23 Gbps on the 256-bit bus interface. This will provide up to 736 GB/s of bandwidth. This is still slightly slower than the 760 GB/s bandwidth offered by the RTX 3080 as it comes with a 320-bit interface but a low 10 GB capacity. To compensate for the lower bandwidth, NVIDIA could integrate a next-generation memory compression suite to replace the 256-bit interface.

For power, the TBP is now set to 340W, a 20W increase from the previous 320W spec we got. This brings TBP to the same ballpark as existing RTX 3080 graphics cards (up to 350W). It is currently unknown if other RTX 40 series graphics cards will also get the faster GDDR6X memory treatment but we do know that Micron has. started full mass production of up to 24 Gbps of GDDR6X memory modules so they have to go somewhere.

  • NVIDIA GeForce RTX 4080 TBP “Expected” – 340W
  • NVIDIA GeForce RTX 3080 “Official” TBP – 350W

NVIDIA GeForce RTX 4080 Series Initial Specifications:

Graphics Card Name NVIDIA GeForce RTX 4080 Ti NVIDIA GeForce RTX 4080 NVIDIA GeForce RTX 3090 Ti NVIDIA GeForce RTX 3080
GPU name Do you have a Lovelace AD102-250? Is there a Lovelace AD103-300? Ampere GA102-225 Ampere GA102-200
Process Node TSMC 4N TSMC 4N Samsung 8nm Samsung 8nm
Die Size ~450mm2 ~450mm2 628.4mm2 628.4mm2
transistor TBD TBD 28 Billion 28 Billion
CUDA Color 14848 9728? 10240 8704
TMU / ROP TBD / 232? TBD/214? 320 / 112 272 / 96
Tensor Core / RT TBD / TBD TBD / TBD 320/80 272 / 68
Base Clock TBD TBD 1365 MHz 1440 MHz
Increase Clock ~2600MHz ~2500MHz 1665 MHz 1710 MHz
Calculate FP32 ~55TFLOP ~50 TFLOP 34 TFLOP 30 TFLOP
Tensor-TOP TBD TBD 273 TOP 238 TOP
Memory capacity 20GB GDDR6X 16GB GDDR6X?
Memory Bus 320-bit 256-bit?
384-bit 320-bit
Memory Speed 21.0 Gbps? 23.0 Gbps? 19 Gbps 19 Gbps
Bandwidth 840 GB/s 736 GB/s?
552 GB/s?
912 Gbps 760 Gbps
TBP 450W 340W 350W 320W
Price (MSRP/FE) $1199 US? $699 US? $1199 $699 US
Launch (Availability) 2023? July 2022? June 3, 2021 17 september 2020

NVIDIA GeForce RTX 4090 Specs ‘Expected’

The NVIDIA GeForce RTX 4090 will use 128 SM from 144 SM for a total of 16,384 CUDA cores. The GPU will be packed with 96 MB of L2 cache and a total of 384 ROP which is absolutely insane. The clock speed is unconfirmed but given the TSMC 4N process in use, we expect the clock to be between the 2.0-3.0 GHz range.

As for memory specifications, the GeForce RTX 4090 is expected to have a GDDR6X capacity of 24 GB which will be clocked at 21 Gbps on the 384-bit bus interface. This will provide up to 1 TB/s bandwidth. This is the same bandwidth as the existing RTX 3090 Ti graphics card and as far as power consumption is concerned, the TBP is said to be rated at 450W which means that the TGP might end up lower than that. The card will be powered by a single 16-pin connector that delivers up to 600W of power. Chances are we could get a custom 500W+ design like the one we saw with the RTX 3090 Ti.

As for the feature sets, the NVIDIA GeForce RTX 4090 and RTX 4080 graphics cards will rock all modern NV feature sets like the latest 4th Gen Tensor Cores, 3rd Generation RT cores, the latest NVENC Encoder and NVCDEC Decoder, and support for the latest APIs. They will pack all modern RTX features like DLSS, Reflex, Broadcast, Resizable-BAR, Freestyle, Ansel, Highlights, Shadowplay, and G-SYNC support as well.

  • NVIDIA GeForce RTX 4090 TBP “Expected” – 450W
  • NVIDIA GeForce RTX 3090 “Official” TBP – 350W

NVIDIA GeForce RTX 4090 Ti & RTX 4090 ‘Startup’ Specs:

Graphics Card Name NVIDIA GeForce RTX 4090 Ti NVIDIA GeForce RTX 4090 NVIDIA GeForce RTX 3090 Ti NVIDIA GeForce RTX 3090
GPU name Do you have a Lovelace AD102-350? Is there a Lovelace AD102-300? Ampere GA102-350 Ampere GA102-300
Process Node TSMC 4N TSMC 4N Samsung 8nm Samsung 8nm
Die Size ~600mm2 ~600mm2 628.4mm2 628.4mm2
transistor TBD TBD 28 Billion 28 Billion
CUDA Color 18432 16128 10752 10496
TMU / ROP TBD / 384 TBD / 384 336 / 112 328 / 112
Tensor Core / RT TBD / TBD TBD / TBD 336 / 84 328 / 82
Base Clock TBD TBD 1560 MHz 1400 MHz
Increase Clock ~2800MHz ~2600MHz 1860 MHz 1700 MHz
Calculate FP32 ~103 TFLOP ~90 TFLOP 40 TFLOP 36 TFLOP
Tensor-TOP TBD TBD 320 TOP 285 TOP
Memory capacity GDDR6X 24GB GDDR6X 24GB GDDR6X 24GB GDDR6X 24GB
Memory Bus 384-bit 384-bit 384-bit 384-bit
Memory Speed 24.0 Gbps 21.0 Gbps 21.0 Gbps 19.5 Gbps
Bandwidth 1152 GB/s 1008GB/s 1008GB/s 936 Gbps
TGP 600W 450W 450W 350W
Price (MSRP/FE) $199 US? $1499 US? $1999 US $1499 US
Launch (Availability) 2023? October 2022? March 29, 2022 24 September 2020

The NVIDIA GeForce RTX 4080 and RTX 4070 graphics cards will be among the first graphics cards other than the RTX 4090 to launch for gamers, but we can’t say for sure if the green team plans to launch them within the year immediately after the RTX 4090 or wait until early next year. The RTX 4090 is so far expected to launch sometime in October 2022.

Which NVIDIA GeForce RTX 40 series graphics card are you most looking forward to?

Products mentioned in this post

By Blanca

Leave a Reply

Your email address will not be published. Required fields are marked *