Tuesday, May 6th 2025
NVIDIA Announces GeForce RTX 5060 Availability Date and Specifications
NVIDIA today announced that the GeForce RTX 5060 (non-Ti) graphics card will be available from May 19, 2025. This includes availability of graphics cards from NVIDIA's board partners, and pre-built desktops with RTX 5060. The RTX 5060 was announced by NVIDIA alongside the RTX 5060 Ti in April, however, availability of the RTX 5060 was pushed to sometime in mid-May. With the 2025 Computex Taipei getting underway on May 20, and the press expected to start their coverage of the expo by May 16 at least; it's highly likely that launch of the RTX 5060 will get obscured by the deluge of Computex content.
NVIDIA also confirmed the specifications of the GeForce RTX 5060. The card is based on the same GB206 silicon as the RTX 5060 Ti, but is configured with 30 streaming multiprocessors (SM) out of the 36 present on the silicon. This works out to 3,840 CUDA cores, 120 Tensor cores, 30 RT cores, 120 TMUs, and 48 ROPs. The GPU is clocked at 2.50 GHz max boost. The RTX 5060 only gets an 8 GB memory option, there's no 16 GB. The 8 GB of GDDR7 memory runs across the GPU's 128-bit wide memory interface. The company didn't mention the memory speed. NVIDIA configured the TGP of the RTX 5060 to be 145 W, a significant reduction from the 180 W of the RTX 5060 Ti, which means most partner cards should make do with a single 8-pin PCIe power connector; it wouldn't surprise us if some cards even come with 6-pin.
Source:
NVIDIA GeForce (Twitter)
NVIDIA also confirmed the specifications of the GeForce RTX 5060. The card is based on the same GB206 silicon as the RTX 5060 Ti, but is configured with 30 streaming multiprocessors (SM) out of the 36 present on the silicon. This works out to 3,840 CUDA cores, 120 Tensor cores, 30 RT cores, 120 TMUs, and 48 ROPs. The GPU is clocked at 2.50 GHz max boost. The RTX 5060 only gets an 8 GB memory option, there's no 16 GB. The 8 GB of GDDR7 memory runs across the GPU's 128-bit wide memory interface. The company didn't mention the memory speed. NVIDIA configured the TGP of the RTX 5060 to be 145 W, a significant reduction from the 180 W of the RTX 5060 Ti, which means most partner cards should make do with a single 8-pin PCIe power connector; it wouldn't surprise us if some cards even come with 6-pin.
15 Comments on NVIDIA Announces GeForce RTX 5060 Availability Date and Specifications
The card will sell. It will sell a lot.
At lower resolutions, where the 8GB VRAM is still somewhat relevant, the vanilla 5060 will probably end up uncomfortably close to the Ti, especially when overclocked.
At least, with the much lower board power, it has a good chance at being a top efficiency card!
I do wonder, though, how far it could be pushed.
Especially if 8-pin connectors, with their massive safety margens, are used, a shunt modded 5060 could probably stand toe to toe with an 8GB Ti.
Something relatively powerful (that isn't Strix Halo) like an 890M is more akin to something like a GTX 1060. It's much better than how iGPUs were back in the day, but let's not lose perspective here.
"High-End" Tier: £400+
AKA "The 'I Accidentally Clicked 'Buy Now' During A Midlife Crisis" Tier
- For products that whisper "premium" but actually mean "you could've bought a functional kidney instead."
- Features include: marginally better numbers, RGB lighting that does nothing, and the crushing realization that your savings account now has trust issues.
"Halo" Tier: £600+AKA "The Unicorn Tears Collection"
- Reserved for items forged in the fires of corporate hubris, like flagship products dipped in liquid ego.
- Example: A graphics card that costs as much as a used Honda Civic but delivers 7% more fps and 100% more buyer's remorse.
- Bonus: Comes with a free aura of smugness (until the next-gen model drops 3 weeks later).
Ultra-Luxury/"Specialized Hi-Fi Financial Ruin" Tier: £800+AKA "The 'I’ve Given Up on Retirement, Let’s YOLO' Tier"
[/HR]
GPU Pricing Rant Addendum:
Re: the Radeon 7800 XT ever dipping below £300…
At this rate? You’ll see pigs piloting fighter jets before GPU prices show mercy. NVIDIA and AMD are too busy inventing new tiers like "The 'We Added a Single Extra Pixel, Give Us Your Paycheck'" Edition. My advice? Start a "GPU Fund" jar now. Or just embrace integrated graphics—potato-powered gaming is retro chic, right?
Stop trashing the thread.
And, stop troll posting.