Just in time for summer, Nvidia’s “Blackwell” graphics cards are coming to the masses, awesome new DLSS 4 features in tow. Today, Nvidia unveiled not one, not two, but three new GeForce RTX 5060-series GPUs, starting at $299.
That $299 offering, the vanilla 8GB GeForce RTX 5060, won’t hit the streets until sometime in May. But the GeForce RTX 5060 Ti arrives tomorrow, April 16, with your choice of memory configurations: 8GB for $379 or 16GB for $429.
While the RTX 5060 Ti has graphical prowess to play at 1440p in many titles, these GPUs will probably be better suited for 1080p gaming given their tiny 128-bit bus and mere 8GB of GDDR7 memory in all but the tippity-top offering. Check out the specs below.
Nvidia
Those paltry 8GB memory buffers are sure to raise some eyebrows. Gamers have been saying “8GB ain’t enough” for years now, as numerous cutting-edge games have run into performance issues caused by exceeding 8GB memory capacity. Think lowered texture details, stuttering, or even a flat-out refusal to run in some cases.
Intel’s $250 Arc B580 packs 12GB of (slower GDDR6) memory across a wider 192-bit bus to avoid capacity issues, but Nvidia decided to stick with the smaller amount, ostensibly to reduce cost.
GeForce product management director Justin Walker acknowledged that 8GB of VRAM capacity might not deliver maximum performance in all circumstances, especially if settings are pushed in intense games. “If you want to play Indiana Jones and the Great Circle maxed out at 1440p resolution, yeah, i think the 16GB version is for you.” That said, Walker asked, “How do we get you a product with the best possible performance for dollar?” He said Nvidia balances by targeting most use cases. Most people playing most games, especially outside of Ultra settings, will do fine with 8GB of VRAM, he said.
RTX 5060 Ti benchmark numbers provided by Nvidia.
Nvidia
RTX 5060 benchmark numbers provided by Nvidia.
Nvidia
To back up that assertion, Nvidia released benchmark charts for the GeForce RTX 5060 Ti and RTX 5060 showing that the cards perform and respond much faster than the RTX 3060 Ti (which doesn’t support AI Frame Generation) and the 4060 Ti (which only supports 2x Frame Gen) when you activate Nvidia’s killer suite of features. Those include normal DLSS upscaling, 4x Multi Frame Generation, and the latency-busting Nvidia Reflex tool. (Delta Force in the charts above lack Frame Gen, but support Reflex.)
That was the theme of Nvidia’s press briefing: Don’t just look at base rasterization numbers (which Walker said would only be 20 to 25 percent faster than the 4060 series) but the peak experience, with all features activated. Not everyone embraces AI features, but a whole lot of gamers do, and the screenshots below show what the RTX 5060 series is capable of if you go whole-hog on Nvidia features.
Nvidia
Nvidia
Nvidia
I said it before, and I’ll say it again: Nvidia bet the entire RTX 50-series on DLSS 4’s Multi Frame Gen feature. Fortunately, it works like black magic, delivering exceptional frame pacing and ultra-smooth visuals with good responsiveness – assuming you’re able to hit roughly a 60 frames-per-second base rendering frame rate before flipping MFG on.
Will the RTX 5060 and 5060 Ti be able to do that? We’ll know when those cards launch as a barrage of custom models — there’s no Founders Edition for these parts — starting tomorrow (April 16) for the 5060 Ti, followed by the RTX 5060 in May. Independent reviews should hopefully give us a full scope of the RTX 5060 series’ capability sooner than later. Stay tuned.
Nvidia