This is simple. If the GTX 1660 outsold the GTX 1650 by 4-to-1 in its launch year, but only 130K units compared to 230K units overall, then the GTX 1650 throttled it in COVID & post-COVID sales, and was therefore a respectable seller during that period. And the 1650 overall sales were obviously strong despite not being nearly as appealing as the 1050, 750, or other >75W cards before. If it "lagged behind", then so did ever other GPU that wasn't the 1660 Super.
The fact that $200-$300 has been the bestselling space for over a decade isn't news to anyone.
They wound down 1660 supply fast in favor of 1660 super and 1660 Ti. You'll see that in sell in modeling from that time period, as well as very clearly in sell out. Again, SI's and stores don't want to be selling 75W GPUs, whether on their own or in prebuilds. The margins and demand weren't there during pandemic, and they are definitely not there now.
LOL, you don't understand the difference, do you? HDDs aren't as demanded because they were supplanted by SSDs which dropped to roughly the same price. Why buy the less for the same? Meanwhile, those office comps are still moving massive units on Amazon, and there is no cheaper a way to transform them into viable gaming comps than with a simple >75W GPU upgrade that can fit into the comp, and run with its PSU.
Office comps don't move massive units. The used channel doesn't even move as much volume as office supply stores, which are already low volume in the grand scheme. The point was just as HDD's were surpassed by the market, so have 75w GPUs. There's just no appetite from them from CyberPower and iBuyPower. And if they won't buy something, a GPU manufacturer has no hope of selling volume.
The average gamer absolutely has no interest in building their own system, let alone repurposing a office supply chassis. You're making ridiculous claims with no actual hard data to back it up. That you think amazon best seller lists and the steam survey are accurate says enough.
Your reading comprehension is clearly the poor one. It's easy to win a market segment that NVIDIA isn't interested in supplying.
No it is not. GTX 1050 gamers haven't been holding out for 75W GPUs. They just bought up, to -60 tier cards or 500/600 tier cards. There aren't millions of gamers that have been waiting to buy, that's not how the mass market works.
You keep ducking because you realize you're in over your head.
What do you think the TAM for 75W GPUs in the US is right now?
Blah blah blah. You're evading the central conundrum. Do you think Battlemage's strategy will work? Do you think these cards are going to sell well?
Either you say yes, or you offer a better alternative strategy.
I think it will provide valuable experience for iGPUs and some datacenter IP. Sales will be insignificant, however, as Intel doesn't have the technological prowess yet to hit 75W performance worth a damn at $150 or $200. GPUs are very hard to design and build, and it's much easier to scale up your architecture than it is to scale down.
Once again, classic MadMick handwaving. You keep insisting that Intel can do 75W GPUs that sell like hotcake even though at almost 200Ws they are only hitting 4060 level performance. There isn't the maturity to hit 75W and 4050 performance without absolutely getting murdered because Intel dies are gigantic still. I'm sure you just have this brilliant plan that for some reason all the brightest minds in the GPU industry couldn't fathom.
