NVIDIA GeForce RTX 4060 Ti “AD106-350” GPU Pictured, Uses Samsung GDDR6 Dies
By: Hassan Mujtaba
Source: https://wccftech.com/nvidia-geforce-rtx-4060-ti-ad106-350-gpu-pictured-uses-samsung-gddr6-dies/
The AD106 Ada Lovelace GPU for NVIDIA’s upcoming GeForce RTX 4060 Ti graphics card has been pictured.
NVIDIA GeForce RTX 4060 Ti GPU Leak Confirms AD106-350 Ada GPU, PCB Shot Shows Use of Samsung GDDR6 Dies
In a picture leaked by MEGAsizeGPU, we get our very first look at the Ada Lovelace GPU that will be powering the NVIDIA GeForce RTX 4060 Ti graphics card. This chip in particular is the “AD106-350” die which was earlier reported to power the graphics card and now this leak more or less confirms this. The PCB also shows the use of Samsung GDDR6 dies which means that lower-tier cards will be making use of a different memory design than what’s featured on the higher-end GDDR6X designs.
4060Ti pic.twitter.com/BVmYyj95g6
— MEGAsizeGPU (@Zed__Wang) April 28, 2023
The use of the AD106 GPU for the GeForce RTX 4060 Ti confirm that it is being made use of a lower-tier chip than the previous-gen. The RTX 3060 Ti featured the GA104 and a 256-bit bus interface but the RTX 4060 Ti is going to make use of the AD106 GPU and a 128-bit bus interface. The Samsung dies are also going to operate at lower speeds than the GDDR6X chips so even if NVIDIA uses the 18 Gbps modules, it will lead to just 288 GB/s bandwidth which will be lower than the 3060 Ti’s 360 GB/s bandwidth.
The one advantage the card will have is a higher L2 cache besides the higher power efficiency. The 8 GB VRAM might be a bit troublesome to make the card attractive to gamers. There have been a large number of unoptimized PC ports over the past few years which don’t run well on cards with lower VRAM and demand at least 10-12 GB VRAM buffer for an adequate gaming experience. How NVIDIA will market the 8 GB on the RTX 4060 Ti graphics card remains to be seen.
NVIDIA GeForce RTX 4060 Ti Specifications “Rumored”:
The GeForce RTX 4060 Ti is expected to utilize the AD106-350-A1 GPU core, a cut-down version of the full AD106 graphics chip, and based on rumors, it should pack 34 SMs or 4352 CUDA cores, an 8 GB GDDR6 memory running at 18 Gbps across a 128-bit bus interface, providing the card with 288 GB/s of bandwidth. There’s also 32 MB of L2 cache on board the GPU which is an 8x increase over the GeForce RTX 3060 Ti.
Using the rumored “reference” boost frequency as a reference, the NVIDIA GeForce RTX 4060 Ti should offer around 22 TFLOPs of compute performance with AIB models boosting it close to 24 TFLOPs.
FP32 Compute Horsepower Comparisons (Higher is Better)
Based on the numbers above, it looks like the NVIDIA GeForce RTX 4060 Ti graphics card is going to end up similar to the GeForce RTX 3070 Ti in performance. In addition to that, the card is going run at a TGP of around 150W which will make it incredibly efficient versus the 3070 Ti while offering faster Raytracing performance with the added bonus of DLSS 3. The RTX 4060 Ti might also bring back ITX designs to the table which makes perfect sense for a chip this efficient.
The NVIDIA GeForce RTX 4060 Ti card is expected to be introduced and launched next month in May. Pricing is expected to be set between $399-$499 US and performance might end up being close to the GeForce RTX 3070 Ti.
NVIDIA GeForce RTX 40 Series “Expected” Lineup Specs:
| Graphics Card Name | NVIDIA GeForce RTX 4090 | NVIDIA GeForce RTX 4080 | NVIDIA GeForce RTX 4070 Ti | NVIDIA GeForce RTX 4070 | NVIDIA GeForce RTX 4060 Ti | NVIDIA GeForce RTX 4060 | NVIDIA GeForce RTX 4050 |
|---|---|---|---|---|---|---|---|
| GPU Name | Ada Lovelace AD102-300(1) | Ada Lovelace AD103-300(1) | Ada Lovelace AD104-400(1) | Ada Lovelace AD104-250(1) | Ada Lovelace AD106-350 | Ada Lovelace AD107-400 | Ada Lovelace AD107 |
| Process Node | TSMC 4N | TSMC 4N | TSMC 4N | TSMC 4N | TSMC 4N | TSMC 4N | TSMC 4N |
| Die Size | 608mm2 | 378.6mm2 | 294.5mm2 | 294.5mm2 | 190.1mm2 | TSMC 4N | TSMC 4N |
| Transistors | 76 Billion | 45.9 Billion | 35.8 Billion | 35.8 Billion | TBD | TBD | TBD |
| CUDA Cores | 16384 | 9728 | 7680 | 5888 | 4352 | 3072 | 2560 |
| TMUs / ROPs | 512 / 176 | 320 / 112 | 240 / 80 | 184 / 64 | TBD | TBD | TBD |
| Tensor / RT Cores | 512 / 128 | 304 / 76 | 240 / 60 | 184 / 46 | TBD | TBD | TBD |
| Base Clock | 2230 MHz | 2210 MHz | 2310 MHz | 1920 MHz | 2310 MHz | TBD | TBD |
| Boost Clock | 2520 MHz | 2510 MHz | 2610 MHz | 2475 MHz | 2535 MHz | TBD | TBD |
| FP32 Compute | 83 TFLOPs | 49 TFLOPs | 40 TFLOPs | 29 TFLOPs | 22 TFLOPs | TBD | TBD |
| RT TFLOPs | 191 TFLOPs | 113 TFLOPs | 82 TFLOPs | TBD | TBD | TBD | TBD |
| Tensor-TOPs | 1321 TOPs | 780 TOPs | 641 TOPs | TBD | TBD | TBD | TBD |
| Memory Capacity | 24 GB GDDR6X | 16 GB GDDR6X | 12 GB GDDR6X | 12 GB GDDR6X | 8 GB GDDR6 | 8 GB GDDR6 | 6 GB GDDR6 |
| Memory Bus | 384-bit | 256-bit | 192-bit | 192-bit | 128-bit | 128-bit | 96-bit |
| Memory Speed | 21.0 Gbps | 22.4 Gbps | 21.0 Gbps | 21.0 Gbps | 18 Gbps | 18 Gbps | TBD |
| Bandwidth | 1008 GB/s | 717 GB/s | 504 GB/s | 504 GB/s | 288 GB/s | 288 GB/s | TBD |
| TBP | 450W | 320W | 285W | 200W | 160W | 115W | ~75W |
| Price (MSRP / FE) | $1599 US / 1949 EU | $1199 US / 1469 EU | $799 US | $599 US | TBD | TBD | TBD |
| Price (Current) | $1599 US / 1859 EU | $1199 US / 1399 EU | $799 US | $599 US | TBD | TBD | TBD |
| Launch (Availability) | 12th October 2022 | 16th November 2022 | 5th January 2023 | 13th April, 2023 | May 2023 | May 2023 | June 2023 |