NVIDIA is working on an RTX 5070 Ti SUPER. It's rumored to arrive in late 2026 or early 2027, though nothing is official yet. The key change is 24GB of memory instead of 16GB. The GPU cores stay the same. The power draw goes up to 350W.
That's basically it. A memory upgrade dressed up as a new product.
Why The Delays
NVIDIA needs special 3GB GDDR7 memory modules for this card. Those modules don't exist in large quantities yet, which is why the SUPER refresh keeps getting pushed back. Memory makers are focused on AI, so consumer GPU memory takes a backseat.
This is the same problem affecting everything else in the market right now. Memory is expensive and hard to find. NVIDIA's stuck waiting like everyone else.
The Root Cause
AI datacenters are consuming 3GB GDDR7 modules faster than memory fabs can produce them. Samsung, SK Hynix, and Micron are prioritizing HBM and high-capacity GDDR for AI workloads over consumer GPU modules. Until that changes, the RTX 5070 Ti SUPER has no guaranteed production slot.
What The RTX 5070 Ti SUPER Actually Offers
Same GPU as the regular RTX 5070 Ti. Same 8,960 CUDA cores. Same clock speeds, probably a bit higher thanks to the extra 50W power budget, but nothing dramatic. Same memory bandwidth of 896 GB/s because the memory bus is still 256-bit.
The only real difference is 24GB versus 16GB. That's a 50 percent capacity bump — and whether that matters depends entirely on what you do with your GPU.
Gaming (1440p / 4K)
24GB won't move the needle. The RTX 5070 Ti already handles high resolutions fine and no current game saturates 16GB. The extra memory will be useful in the future though.
Not worth it for gamingCreative & AI Workloads
Video editing, 3D rendering, AI image generation, and local LLM inference can all exhaust 16GB fast. 24GB is the difference between your software running clean or running compromised.
Genuinely useful herePerformance Expectations
Gaming performance should be roughly the same as the regular RTX 5070 Ti, maybe a bit higher if NVIDIA can push the clocks with the extra 50W headroom. Nothing significant. If you're buying this for gaming, the extra VRAM is not worth the wait or the price.
For creative work or AI inference, the extra VRAM is actually useful — and that's clearly the target audience NVIDIA is designing this card for.
+50%
VRAM increase vs standard RTX 5070 Ti
+50W
Higher TDP (350W vs 300W)
Price And Timeline
NVIDIA hasn't said anything official. The regular RTX 5070 Ti launched at $749. The SUPER version will probably cost more because memory is expensive and 3GB GDDR7 modules cost more per gigabyte than the standard 2GB modules used in the base card. How much more is impossible to predict until NVIDIA actually announces it.
The launch has been pushed back repeatedly since rumors started in April 2025. Right now the best guess is early 2027, but that could change if memory supply improves — or get pushed further if GDDR7 supply tightens even more.
Bottom Line
This card exists in rumor form. It'll have 24GB of memory. It'll probably cost more than $749. It might launch in 2027. That's all anyone actually knows right now.
Timeline of RTX 5070 Ti SUPER Rumors
Over a year of delays. The pattern is clear: memory shortage → delay → memory shortage again → new delay. This will probably keep happening until GDDR7 supply normalizes or NVIDIA decides to cancel the whole thing.
Thirteen separate leaks or reports over fourteen months. Zero official announcements. That's the RTX 5070 Ti SUPER story so far.
Share this article to fellow gamers
