Monday, December 16th 2024
NVIDIA GeForce RTX 5070 Ti Leak Tips More VRAM, Cores, and Power Draw
It's an open secret by now that NVIDIA's GeForce RTX 5000 series GPUs are on the way, with an early 2025 launch on the cards. Now, preliminary details about the RTX 5070 Ti have leaked, revealing an increase in both VRAM and TDP and suggesting that the new upper mid-range GPU will finally address the increased VRAM demand from modern games. According to the leak from Wccftech, the RTX 5070 Ti will have 16 GB of GDDR7 VRAM, up from 12 GB on the RTX 4070 Ti, as we previously speculated. Also confirming previous leaks, the new sources confirm that the 5070 Ti will use the cut-down GB203 chip, although the new leak points to a significantly higher TBP of 350 W. The new memory configuration will supposedly run on a 256-bit memory bus and run at 28 Gbps for a total memory bandwidth of 896 GB/s, which is a significant boost over the RTX 4070 Ti.
Supposedly, the RTX 5070 Ti will also see a bump in total CUDA cores, from 7680 in the RTX 4070 Ti to 8960 in the RTX 5070 Ti. The new RTX 5070 Ti will also switch to the 12V-2x6 power connector, compared to the 16-pin connector from the 4070 Ti. NVIDIA is expected to announce the RTX 5000 series graphics cards at CES 2025 in early January, but the RTX 5070 Ti will supposedly be the third card in the 5000-series launch cycle. That said, leaks suggest that the 5070 Ti will still launch in Q1 2025, meaning we may see an indication of specs at CES 2025, although pricing is still unclear.
Update Dec 16th: Kopite7kimi, ubiquitous hardware leaker, has since responded to the RTX 5070 Ti leaks, stating that 350 W may be on the higher end for the RTX 5070 Ti: "...the latest data shows 285W. However, 350W is also one of the configs." This could mean that a TBP of 350 W is possible, although maybe only on certain graphics card models, if competition is strong, or in certain boost scenarios.
Sources:
Wccftech, Kopite7kimi on X
Supposedly, the RTX 5070 Ti will also see a bump in total CUDA cores, from 7680 in the RTX 4070 Ti to 8960 in the RTX 5070 Ti. The new RTX 5070 Ti will also switch to the 12V-2x6 power connector, compared to the 16-pin connector from the 4070 Ti. NVIDIA is expected to announce the RTX 5000 series graphics cards at CES 2025 in early January, but the RTX 5070 Ti will supposedly be the third card in the 5000-series launch cycle. That said, leaks suggest that the 5070 Ti will still launch in Q1 2025, meaning we may see an indication of specs at CES 2025, although pricing is still unclear.
Update Dec 16th: Kopite7kimi, ubiquitous hardware leaker, has since responded to the RTX 5070 Ti leaks, stating that 350 W may be on the higher end for the RTX 5070 Ti: "...the latest data shows 285W. However, 350W is also one of the configs." This could mean that a TBP of 350 W is possible, although maybe only on certain graphics card models, if competition is strong, or in certain boost scenarios.
161 Comments on NVIDIA GeForce RTX 5070 Ti Leak Tips More VRAM, Cores, and Power Draw
The 4070Ti Super is already a tad slower than the 4080.
Remember that the 40 series cards actually had lower bandwidth than 30 series: Nvidia slashed bus width and only slightly increased memory clock rate, even as they raised core counts and core clocks by a huge amount. Now 50 series is going to barely improve core counts, but increase bandwidth 30-50% from the previous generation?!
Can anyone weigh in: How would it benefit gaming? Seems mostly a boon for LLM inference speed.
Edit: and a big increase in TDP without increase in core count probably means throwing efficiency out the window to crank clocks. Although leaked TDPs in the past have been way off: kopite7kimi said the 4080 would have 420W TDP, but it launched with 320W TDP.
Nvidia tried to justify the high price of the 4070 Ti last time by calling it "4080 12GB".
Now with no competition and the public hypnotized by Nvidia's logo, they can easily put a $900 price tag and still sell it like hot cakes. If they come out with a $800 price tag, the original price tag for the 4070 Ti, the hordes will go bananas, looking at Jensen like he is the biggest humanitarian alive. The few who will think that this is a high unrealistic price, will just blame AMD and start praying for Intel to save them..... in 5 years.They will probably support a new feature that will be justifying an upgrade. Think Frame Generation as the main difference between RTX 3060 and RTX 4060. A new feature for the RTX 5000 series that could be supported by RTX 4000 and probably even RTX 3000 series but Nvidia will say that "it can't be supported on older GPUs because of hardware changes in RTX 5000 that are necessary". Then AMD and Intel will introduce that feature for their cards, making it obvious that Nvidia lies, but I doubt they will give it to Nvidia owners this time.42% faster (based on TPU) with a 33% higher MSRP.Nvidia doesn't sell only performance. It also sells the feeling of having a superior product compared to the others. Think Apple vs Samsung. Even if Samsung comes out with a superior model, people will still think that buying an Apple iPhone is the premium choice.
Or that previous post where you where saying that many people "just hate Nvidia".
Or maybe we should just combine those two posts.
"Many people hate Nvidia, most posters here have an AMD GPU. lol".
How about that?
..........please.......... The above two posts of yours could be considered too close to an effort to ignite a flame war.
Peace and thermal grease y0
Let's hope some semblance of sanity has been restored with pricing, as it only took 1 generation to see a doubling of prices for xx70 cards from $400 to over $600+
Anyways, time will tell.
The results:
1. No future proofing
2. Reduced textures resolution in order to fit them in the limited framebuffer
See this for a reference at 6:06, 8:03:
GTX 1070 Ti was a 180W card
www.techpowerup.com/gpu-specs/geforce-gtx-1070-ti.c3010
Question: why did they backport GB102 on 3nm to GB202 on 4nm? Does it mean the TSMC 3nm node is broken for GPUs ? :rolleyes:
24GB on a 5080 should now be a thing nGreedia, as it should have been all along. But I won't hold my breath.
But I'm not impressed by 16GB on the 5070Ti which will probably be $800-900, its the minimum a card should have at that price.I don't expect anyone to feel like they're getting screwed while still having any excitement for new GPU's from Nvidia, and I mean anyone still accepting what the leather jacket man is charging for what you get in a mid range card.
As for inflation, it doesn't really apply when Nvidia has margins over 70% and they have 90% of the dGPU market, they're pricing things well above inflation because they can.