This past January at CES Nvidia announced their new 50 Series of GPUs based on their Blackwell architecture. The flagship of the new lineup was their RTX 5090 but they did also announce a few of the models below it including the RTX 5080, RTX 5070 Ti, and RTX 5070. I covered the RTX 5090 and the RTX 5080 late in January but now we start to get into the other models with the RTX 5070 Ti embargo lifting today and RTX 5070 Ti’s hitting stores tomorrow. The RTX 5070 Ti isn’t available as a Founders Edition card, similarly to the RTX 4070 Ti so for those considering getting an RTX 5070 Ti, I have the Asus Prime RTX 5070 Ti on hand to check out its performance. So let’s dive in and see what it's all about.
Product Name: Asus Prime RTX 5070 Ti
Review Sample Provided by: Asus
Written by: Wes Compton
Amazon Affiliate Link: HERE
What is new?
Nvidia announced a lot with this one so let’s try to touch on as much as we can. They of course had a few hardware announcements for the 50-Series. They have announced four cards, the RTX 5090, the RTX 5080, the RTX 5070 Ti, and the RTX 5070. That is the new 50-series RTX family. The slides below include their focus on them. The 5090 for example is targeted at 4k 240 Hz and the other three are targeted at 2k or 1440p performance. They all have GDDR7 VRAM with the flagship RTX 5090 having 32 GB and a 512-bit interface. The RTX 5080 cuts that in half to 16 GB, the RTX 5070 Ti also has 16GB, and then the RTX 5070 has 12 GB.
The new GPUs are based on the Blackwell architecture and they do have new updated RT and Tensor cores with that being the 4th gen for the RT Cores and 5th gen for Tensor cores. The new tensor cores can now also handle floating point (FP) 4 along with FP8 and FP16. They also have introduced an AMP processer which is the AI Management Processor to help schedule AI tasks alongside of graphics rendering. They have increased the number of NV encoders and decoders, it now depends on the card model and isn’t a flat number of them across the entire generation of cards. They have also moved to PCIe Gen 5 and DisplayPort 2.1b including UHBR20. For pricing the RTX 5090 comes in at $1999, the RTX 5090 is half that (just like its memory) at $999. The RTX 5070 Ti is $749 and the RTX 5070 is $549.
Some of the main goals with Blackwell were to optimize the neural workloads and reduce the memory footprint. No big surprises there. Memory is one of the more expensive parts of the cards and anyone who has been paying attention knows that neural workloads and AI have been where Nvidia has been seeing the biggest improvements. Combining those things they have brought AI into shaders with neural shaders. The example of this they show is a hair being rendered with ray tracing and how using spheres rather than triangles helps use less data which means less VRAM and higher frame rates.
Nvidia announced Transformer which replaces CNN as the model they use with Super Resolution. They have improved the design significantly to get better detail when scaling this up. On top of that, they announced DLSS 4 which has improved on the frame generation that they introduced with DLSS 3. It is now Multi Frame Generation. Where before they were able to generate every other frame, they can now do x3 or x4. This gets interesting once they add in DLSS Super Resolution as well because that is already rendering ¾ of the image. With both, they are rendering 15 of 16 pixels using AI. Of course how well this works then depends a lot on how good the renders are. But it gives huge improvements in performance and with frame generation, we know that those improvements still happen even if you are PCU limited for example. The example they show has DLSS off at 27 FPS, turning on Super Resolution it goes to 71 FPS. DLSS 3.5 gets you to 140 FPS and DLSS 4 is 248 FPS. DLSS is already supported by a LOT of games and Nvidia is saying that DLSS 4 will have 75 games and apps supporting DLSS 4 at Day 0. Some of those will be by using the Nvidia App, which can override the DLSS settings on some games. It is also important to note that some of these features will go back and work with legacy cards as well. Specifically, DLAA is going to work back to the 20 series of cards, same with the improved DLSS Super Resolution. The new multi-frame generation however only works with 50 series cards.
They also introduced a new Nvidia Reflex. Reflex helps improve responsiveness to get lower latency through the entire pipeline. Were Reflex offered 50% faster responsiveness the new Reflex 2 gets you 75% by using frame warp. They say that is coming first to 50 series and will be available in games like Valorant soon.
Specifications |
RTX 3070 Ti |
RTX 4070 Ti |
RTX 5070 Ti |
RTX 5080 |
GPU Codename |
GA104 |
AD104 |
GB203 |
GB203 |
GPU Architecture |
NVIDIA Ampere |
NVIDIA Ada Lovelace |
NVIDIA Blackwell |
NVIDIA Blackwell |
GPCs |
6 |
5 |
6 |
7 |
TPCs |
24 |
30 |
35 |
42 |
SMs |
48 |
60 |
70 |
84 |
CUDA Cores / SM |
128 |
128 |
128 |
128 |
CUDA Cores / GPU |
6144 |
7680 |
8960 |
10752 |
Tensor Cores / SM |
4 (3rd Gen) |
4 (4th Gen) |
4 (5th Gen) |
4 (5th Gen) |
Tensor Cores / GPU |
192 (3rd Gen) |
240 (4th Gen) |
280 (5th Gen) |
336 (5th Gen) |
GPU Boost Clock (MHz) |
1770 |
2610 |
2452 |
2617 |
RT Cores |
48 (2nd Gen) |
60 (3rd Gen) |
70 (4th Gen) |
84 (4th Gen) |
RT TFLOPS |
42.5 |
92.7 |
133.2 |
170.6 |
Frame Buffer Memory Size and Type |
8 GB GDDR6X |
12 GB GDDR6X |
16 GB GDDR7 |
16 GB GDDR7 |
Memory Interface |
256-bit |
192-bit |
256-bit |
256-bit |
Memory Clock (Data Rate) |
19 Gbps |
21 Gbps |
28 Gbps |
30 Gbps |
Memory Bandwidth |
608 GB/sec |
504 GB/sec |
896 GB/sec |
960 GB/sec |
ROPs |
96 |
80 |
96 |
112 |
Pixel Fill-rate (Gigapixels/sec) |
169.9 |
208.8 |
235.4 |
293.1 |
Texture Units |
192 |
240 |
280 |
336 |
Texel Fill-rate (Gigatexels/sec) |
339.84 |
626.4 |
686.6 |
879.3 |
L1 Data Cache/Shared Memory |
6144 KB |
7680 KB |
8960 KB |
10752 KB |
L2 Cache Size |
4096 KB |
49152 KB |
49152 KB |
65536 KB |
Video Engines |
1 x NVENC (7th Gen) 1 x NVDEC (5th Gen) |
2 x NVENC (8th Gen) 1 x NVDEC (5th Gen) |
2 x NVENC (9th Gen) 1 x NVDEC (6th Gen) |
2 x NVENC (9th Gen) 2 x NVDEC (6th Gen) |
TGP (Total Graphics Power) |
290 W |
285 W |
300 W |
360 W |
Transistor Count |
17.4 Billion |
35.8 Billion |
45.6 Billion |
45.6 Billion |
Die Size |
392.5 mm² |
294.5 mm² |
378 mm² |
378 mm² |
Manufacturing Process |
Samsung 8 nm 8N NVIDIA Custom Process |
TSMC 4nm 4N NVIDIA Custom Process |
TSMC 4nm 4N NVIDIA Custom Process |
TSMC 4nm 4N NVIDIA Custom Process |
PCI Express Interface |
Gen 4 |
Gen 4 |
Gen 5 |
Gen 5 |
Launch MSRP |
$599 |
$799 |
$749 |
$999 |
The RTX 5070 Ti is running the same GB203 GPU as we saw with the RTX 5080 but the 5070 Ti isn’t utilizing the full GPU. When it comes to the GPCs the 5080 had 7 and the 5070 Ti has 6. TPCs go down from 42 to 35 and SMs go from 84 to 70. This is a little more than cutting out one GPC, if that was the case the SM count would be at 72, not 70. The CUDA core count per SM is still the same though and with 14 fewer SMs the RTX 5070 Ti has 8960 CUDA cores to the 10752 of the RTX 5080, a 16% drop. This is still a step up from the 7680 CUDA cores on the RTX 4070 Ti and even more compared to the RTX 3070 Ti which is more likely to be where people might be considering an upgrade. For the Tensor cores the RTX 5070 Ti has 280 to the 336 of the 5080. For comparison, the RTX 4070 Ti had 240 and the 3070 T had 192, that doesn’t also figure in the architecture improvements as well where the 3070 Ti has 3rd Gen Tensor cores, the 4070 Ti had 4th Gen and we are now on the 5th Gen. We see a similar thing with the ray tracing cores as well, the RTX 5070 Ti has 70 to the 84 of the RTX 5080 on the 4th Generation. The RTX 4070 Ti had 60 of the 3rd gen RT cores and the RTX 3070 Ti had 48 of the 2nd gen RT cores. You can see with Nvidia’s RTX TFLOPS numbers how big of an improvement that translates to going from 42.5 on the 3070 Ti up to 92.7 on the 4070 Ti and 133.2 on the new 5070 Ti.
For memory the RTX 5070 Ti has the same 16GB GDDR7 that the RTX 5080 has, this is a step up from the 12GB of GDDR6X on the RTX 4070 Ti and 8 GB of GDDR6X from the RTX 3070 Ti. The memory interface has been bumped back up as well from 192-bit up to 256-bit, just like what we saw on the RTX 5080. The memory bandwidth numbers show just how much that helped going from 504 GB/sec on the RTX 4070 Ti up to 896 GB/sec, the RTX 3070 Ti was a little better with its 256-bit interface but still, it’s a good step up. The base clock for the RTX 5070 Ti is 2452 MHz which is a little lower than the 2617 MHz of the RTX 5080. The RTX 5070 Ti has the upgraded NVENC and NVDEC video engines but there is just one NVDEC where the RTX 5080 has two. The RTX 5070 Ti has a TGP of 300 watts, 60 less than the RTX 5080 and 10 and 15 watts higher than the RTX 3070 Ti and RTX 4070 Ti before it. Like with the other 50 Series GPUs, it is running on PCI Gen 5. Lastly, for pricing, the RTX 5070 Ti has an MSRP of $749, which is $50 less than the launch price of the RTX 4070 Ti but $150 more than the RTX 3070 Ti. Of course that pricing is for the stock-clocked MSRP-focused cards which can be hard or impossible to find post-launch.
Before getting into testing I did also run GPUz to double-check that our clock speeds matched up with the specifications. The Prime RTX 5070 Ti had a clock speed of 2452 MHz which is in line with Nvidia’s stock specifications. I tested using the Nvidia 572.43 driver that Nvidia provided ahead of the launch. And our BIOS version is documented here just in case it needs to be referenced in the future.