The alleged specifications of the NVIDIA GeForce RTX 4070 Ti graphics card have been revealed and it may only match the RTX 3090 Ti in performance.
NVIDIA GeForce RTX 4070 Ti is claimed to feature 7,680 cores, 12GB GDDR6X memory, and RTX 3090 Ti performance
According to the passenger Kopite7kimiHowever, the NVIDIA GeForce RTX 4070 Ti could turn out to be a card monster, delivering performance that matches the GeForce RTX 4090 Ti, the current flagship. The leaker reported some new specifications and it appears that 4070 . series They will benefit from much higher energy limits than their predecessors.
As I mentioned before, the AD104 SKU has a 400W limit.
PG141-SKU331
AD104 Full Fat With 7680FP32
21Gbps 12GDDR6X
It can easily match with RTX 3090 Ti.– kopite7kimi (@kopite7kimi) August 1, 2022
“Expected” NVIDIA GeForce RTX 4070 Specifications
The NVIDIA GeForce RTX 4070 Ti is expected to use a full-fat AD104 GPU core with 7,680 cores, or 60 SM cores. From previous leaks, we also know that the AD104 GPU will come packed with 48MB of L2 cache and up to 160 ROP which is simply insane. That’s a 25% increase in core count and a 12x cache increase versus the GeForce RTX 3070 Ti rocking two GA104 GPU cores.
Clock speeds have yet to be confirmed, but given that the TSMC 4N process is in use, we would expect clocks in the 2.0-3.0GHz range. The higher-than-normal clock speed comes from the fact that NVIDIA is making a two-node jump considering that Ampere GPUs with Samsung’s 8nm node were actually a 10nm process node with some improvements. NVIDIA skips 7nm and goes straight to the 5nm node not even the vanilla variant but an enhanced version of it. With Pascal on the TSMC 16nm node, NVIDIA delivered a massive jump in frequency and we can expect a similar jump this time around as well.
In terms of memory specs, the GeForce RTX 4070 Ti is expected to outperform 12GB GDDR6X which is said to be clocked at 21Gbps via a 192-bit bus interface for 504GB/s bandwidth. While the RTX 4070 is expected to swing around TBP 300 W The RTX 4070 Ti can end up with a power limit of 400 watts. Now, that might be the power limit for custom designs with the Founders Edition roaming around the 325-350W power limit. The graphics card is expected to be based on the PG141-SKU331 PCB design.
As for its feature set, the NVIDIA GeForce RTX 4070 Ti graphics cards will rock all the modern NV feature sets like the latest 4th generation Tensor Cores, 3rd generation RT cores, the latest NVENC Encoder and NVCDEC Decoder, and support for the latest APIs. They’ll pack all the modern RTX features like DLSS, Reflex, Broadcast, Resizable-BAR, Freestyle, Ansel, Highlights, Shadowplay, and G-SYNC support as well. The performance is said to easily match the RTX 3090 Ti which is a $199 card (MSRP). So we might look at that kind of performance on a $599-$649 US graphics card.
NVIDIA GeForce RTX 4070 Series Initial Specifications:
Graphics card name | NVIDIA GeForce RTX 4070 Ti . Graphics Card | NVIDIA GeForce RTX 4070 | NVIDIA GeForce RTX 3070 Ti | NVIDIA GeForce RTX 3070 |
---|---|---|---|---|
GPU name | AD104-400? | AD104-300? | GA104-400 . ampere | GA104-300 . ampere |
Operation knot | TSMC 4N | TSMC 4N | Samsung 8nm | Samsung 8nm |
die size | ~ 300 mm 2 | ~ 300 mm 2 | 395.2 mm 2 | 395.2 mm 2 |
transistors | to be announced later on | to be announced later on | 17.4 billion | 17.4 billion |
PCB | Nvidia PG141-SKU331 | Nvidia PG141-310 SKU341 | Nvidia PG141 | Nvidia PG142 |
CUDA cores | ~ 7680 | ~ 7040 | 6144 | 5888 |
TMUs / ROPs | TBD / 160 | TBD / 144 | 192/96 | 184/96 |
tensor/RT cores | TBD / TBD | TBD / TBD | 192/48 | 184/46 |
base clock | to be announced later on | to be announced later on | 1575 MHz | 1500 MHz |
increase the clock | to be announced later on | to be announced later on | 1770 MHz | 1730 MHz |
FP32 account | ~38 TFLOPs | ~36 TFLOPs | 22 TFLOPs | 20 TFLOPs |
RT TFLOPs | to be announced later on | to be announced later on | 42 TFLOPs | 40 TFLOPs |
Tensor-TOPs | to be announced later on | to be announced later on | 174 up | 163 peak |
Memory capacity | 12 GB GDDR6X? | 12 GB GDDR6 | 8 GB GDDR6X | 8 GB GDDR6 |
Memory bus | 192 bit | 192 bit | 256 bit | 256 bit |
Memory speed | 21 Gbps | 18 Gbps | 19 Gbps | 14 Gbps |
Bandwidth | 504 GB/sec | 432 GB/sec | 608 Gbps | 448 Gbps |
TGP | ~ 400 watts | ~ 300 watts | 290 W | 220 watts |
Price (MSRP/FE) | 599 USD? | 499 US dollars? | 599 USD | 499 USD |
launch (availability) | 2022 | 2022 | June 10, 2021 | October 29, 2020 |
Products mentioned in this post
“Lifelong beer expert. General travel enthusiast. Social media buff. Zombie maven. Communicator.”