Still nothing on prices. I am curious how Blackwell GPUs’ pricing will compare to Intel’s $250 B580.
I’m gonna guess they price the 5600ti at $600.
They could be extremely cheeky and price it at $500, and drop the 4060ti to $400, right after the Xmas shopping season…but I’d guess they don’t even bother to do that.
Don’t think they will be dropping 4060ti prices and inventory for the 4000 series seems to be running out.
It’s nvidia, so it’s gonna be expensive most likely. I hope AMD will have some high vram options.
They will definitely have high VRAM options, but I suspect they will offer a $50 discount (at most) relative to an equivalent Nvidia SKU.
Where does your optimism come from? So far both nvidia and AMD have been very conservative with any sort of vram increases over the past generations. Intel pushes it a little bit within their lower end area but that’s no guarantee the others will follow this trend with their cards.
Just speculation on my part, AMD does tend to offer higher VRAM capacities than Nvidia.
I could be wrong if course.
Triple. Take it or leave it
… Best I can do is tree fiddy.
It was already funny when the 4060 shipped with 8GB
If they’re not willing to make changes to the memory implementation on the 5060 this doesn’t seem that unreasonable tbh. 10-12 with all else being equal on the 5060 would be nice, of course, but $X and 8GB is probably a better market fit than $X+40ish (?) and 16GB.
But that also might be a slightly too convenient way to justify not just tweaking the memory configuration to be a better fit for >8GB but <16GB, so idk lol. I don’t keep up as closely with low level details anymore (e.g., what memory chips are commonly available right now), so it’s hard for me to crucify or defend the choice to avoid a memory config change with much conviction.
They’re jumping straight from 8 to 16. I wonder why they explore the in-betweens so little.
The VRAM size is tied to the memory interface. Basically you can’t just pick a random VRAM amount (short of literally disabling part of it).
Easier to sell the more expensive 16 GB cards. If the difference between the two would be smaller then people may be less inclined to go for the higher model.
Hmm, I see. I imagine there might be an engineering reason too. Maybe there’s some advantage to only ever halving/doubling memory for a given chip?
I guess that could be depending on the producer of those vram chips themselves. If 8 GB is the standard size, then a 12 GB design would be likely more expensive to produce. 24 GB cards could be both out of 2x 12 GB or simply 3x 8 GB, so if you just order a buttload of 8 GB you’d save quite a bit of money.
