Nvidia’s RTX 5080 graphics card will supposedly be loaded with quicker video reminiscence than anticipated, in a brand new twist within the story of rumors being spun round Group Inexperienced’s next-gen GPU.
Wccftech has heard from sources (have interaction skeptical mode now) who declare that the RTX 5080 is ready to be outfitted with 16GB of GDDR7 VRAM (as beforehand rumored), however that this RAM shall be working at 32Gbps, a quicker velocity than the 28Gbps reminiscence that hypothesis prompt prior to now. (Samsung has 28Gbps and 32Gbps GDDR7 modules obtainable initially, as it’s possible you’ll recall).
With a 256-bit reminiscence bus and 32Gbps VRAM, as Wccftech factors out, the full bandwidth on this case will (simply) exceed the magic 1TB/s mark (1024GB/s to be exact). To place that in perspective, the RTX 4080 affords a complete reminiscence bandwidth of 736GB/s.
One other fascinating wrinkle right here is the assertion that the RTX 5090 is ready to make use of 28Gbps reminiscence, nevertheless it’ll have a means wider reminiscence bus at 512-bit, that means its complete bandwidth will nonetheless far exceed the RTX 5080 as you’d count on (it’ll be closing on double, in truth – plus the 5090 will supposedly provide 32GB of VRAM).
Finally, the chances are high that the RTX 5090 may even use 32Gbps video RAM, however initially, that is going to be the area of the RTX 5080 – assuming this information from the grapevine is appropriate (and all the opposite previous rumored specs, for that matter – add seasoning as ever).
Evaluation: Reminiscence issues – some compensation from Nvidia?
Okay, so assuming that is on the cash, why would Nvidia make this VRAM velocity selection for the RTX 5080, then? Effectively, in case you didn’t discover, there’s already been fairly the response to the (purported) reminiscence loadout with the RTX 5080. Firstly, with the selection to make use of that 256-bit bus (sticking with the identical as on the RTX 4080, and never upgrading there), and extra lately, the 16GB of VRAM configuration has left many PC players displeased.
Nvidia doubtless anticipated this, and so maybe that is Group Inexperienced’s means of compensating for these shortcomings to some extent, and making certain general reminiscence bandwidth is suitably peppy. At 1TB/s, or simply over, you’ll be able to’t actually complain – this represents an virtually 40% generational uplift for video reminiscence velocity.
One other optimistic little bit of chatter is that whereas the 16GB loadout of RAM has been extensively seen as Nvidia cheaping out on video reminiscence – as soon as once more – there might be a 24GB model of the RTX 5080 launched down the road.
Worries nonetheless stay in regards to the RTX 5080 general, although, with its rumored core rely representing an enormous drop in comparison with the RTX 5090. The previous could have fewer than half the CUDA Cores of the next-gen flagship, supposedly, a bigger gulf than seen within the RTX 4080 pitted in opposition to the RTX 4090 (the place the distinction is extra like 40%, relatively than 50%).
Are we going to get a nerfed xx80 sequence graphics card for the Blackwell technology? That’s what some are suggesting, however in all honesty, such hypothesis is relatively meaningless till we all know what Nvidia’s pricing plans are for the RTX 5080. Is the thought to angle the GPU a bit extra affordably than the RTX 4080?
The difficulty is, although, it’s troublesome to envisage Group Inexperienced pinning a extra cheap price ticket on a 5080 given previous kind – it’s simply appears like that is severely wishful considering. Nonetheless, all we are able to do is wait and see. In concept the RTX 5090 and 5080 might be unleashed collectively, with an preliminary revelation at CES 2025 (we would even see Blackwell laptop computer GPUs, too).
