A recent driver leak has revealed that NVIDIA is preparing a range of Ada RTX GPUs for the mobile workstation segment.
NVIDIA Has Ada RTX GPUs On The Horizon For Workstation Laptops, RTX 5000 & More Spotted
The NVIDIA RTX Ada Workstation GPUs were spotted within a driver published by MSI which offers some insight into future products and device IDs for GPUs that haven't been released. There are a total of four SKUs that were listed within the driver package which include:
NVIDIA_DEV.2730 = “NVIDIA RTX 5000 Ada Generation Laptop GPU”NVIDIA_DEV.27BB = “NVIDIA RTX 3500 Ada Generation Laptop GPU”NVIDIA_DEV.2838 = “NVIDIA RTX 3000 Ada Generation Laptop GPU”NVIDIA_DEV.28B8 = “NVIDIA RTX 2000 Ada Generation Laptop GPU”
This points out to the next-gen RTX Workstation lineup for laptops utilizing the Ada Lovelace GPU architecture. Now there's no information available besides the Device IDs but Videocardz has the specs for these chips speculated. The tech outlet reports that the RTX 5000 Ada GPU might utilize the AD103 GPU with a 256-bit bus interface and a 16 GB GDDR6 memory buffer, the RTX 3500 can take the AD104 and a 192-bit wide bus interface with a total of 12 GB memory buffer.

On the low-end side, the RTX 3000 and the RTX 2000 Ada Mobility workstation GPUs might utilize the AD106 and AD107 GPUs with an 8 GB 128-bit and 4/8 GB 128-bit memory config, respectively. Their specs may also be similar to the standard RTX 40 Laptop GPUs but with lower clocks and a lower power limit. That's all that we can expect but besides that, the RTX Ada workstation GPUs will definitely take advantage of the Studio driver suite and deliver a nice performance boost over the existing Ampere GPUs.
Considering that there are multiple events from NVIDIA in the coming months such as GDC and GTC, we might hear about their new Ada Workstation GPUs during one of those events.
NVIDIA RTX Workstation GPUs:
| GPU Name | GPU Proocess | GPU SKU | Cores | Tensor Cores | Single-Precision | Memory | Memory Bandwidth | TDP |
|---|---|---|---|---|---|---|---|---|
| RTX 5000 Ada | TSMC 5nm | AD103? Ada | TBD | TBD | TBD | 16 GB GDDR6 (256-bit) | TBD | TBD |
| RTX 3500 Ada | TSMC 5nm | AD104? Ada | TBD | TBD | TBD | 12 GB GDDR6 (192-bit) | TBD | TBD |
| RTX 3000 Ada | TSMC 5nm | AD106? Ada | TBD | TBD | TBD | 8 GB GDDR6 (128-bit) | TBD | TBD |
| RTX 2000 Ada | TSMC 5nm | AD107? Ada | TBD | TBD | TBD | 8 GB GDDR6 (128-bit) | TBD | TBD |
| RTX A5500 | Samsung 8nm | GA103 Ampere | 7424 | 232 | 22.2 TFLOPs | 16 GB GDDR6 (256-bit) | 512 GB/s | 165W |
| RTX A5000 | Samsung 8nm | GA104 Ampere | 6144 | 192 | 19.3 TFLOPs | 16 GB GDDR6 (256-bit) | 448 GB/s | 150W |
| RTX A4500 | Samsung 8nm | GA104 Ampere | 5888 | 184 | 17.6 TFLOPs | 16 GB GDDR6 (256-bit) | 512 GB/s | 140W |
| RTX A4000 | Samsung 8nm | GA104 Ampere | 5120 | 160 | 17.2 TFLOPs | 8 GB GDDR6 (256-bit) | 384 GB/s | 115W |
| RTX A3000 | Samsung 8nm | GA104 Ampere | 4096 | 128 | 10.0 TFLOPs | 6 GB GDDR6 (192-bit) | 264 GB/s | 70W |
| RTX A2000 | Samsung 8nm | GA106 Ampere | 2560 | 80 | 8.63 TFLOPs | 4 GB GDDR6 (128-bit) | 192 GB/s | 95W |
| RTX A1000 | Samsung 8nm | GA107 Ampere | 2048 | 64 | 4.66 TFLOPs | 4 GB GDDR6 (128-bit) | 176 GB/s | 60W |
| RTX A500 | Samsung 8nm | GA107 Ampere | 2048 | 64 | 6.29 TFLOPs | 4 GB GDDR6 (64-bit) | 96 GB/s | 40W |









