Put 32GB on that card and everybody would ignore performance issues.
With 16GB everybody will just call it another in the long list of Intel failures.
With 16GB everybody will just call it another in the long list of Intel failures.
My first software job was at a place doing municipal architecture. The modelers had and needed high end GPUs in addition to the render farm, but plenty of roles at the company simply needed anything with better than what the Intel integrated graphics of the time could produce in order to open the large detailed models.
In these roles the types of work would include things like seeing where every pipe, wire, and plenum for a specific utility or service was in order to plan work between a central plant and a specific room. Stuff like that doesn’t need high amounts of VRAM since streaming textures in worked fine. A little lag never hurt anyone here as the software would simply drop detail until it caught up. Everything was pre-rendered so it didn’t need large amounts of power to display things. What did matter was having the grunt to handle a lot of content and do it across three to six displays.
Today I’m guessing the integrated chips could handle it fine but even my 13900K’s GPU only does DisplayPort 1.4 and up to only three displays on my motherboard. It should do four but it’s up to the ODMs at that point.
For a while Matrox owned a great big slice of this space but eventually everyone fell to the wayside except NVidia and AMD.
Prices from NewEgg on 16gb+ consumer cards, sold by NewEgg and in stock.
Given the high demand of graphic cards, is this a plausible scenario?
Given how young and volatile this domain still is, it doesn't seem unreasonable to be wary of it. Big players (google, openai and the likes) are probably pouring tons of money into trying to do exactly that
But intel is still lost in it's hubris, and still thinks it's a serious player and "one of the boys", so it doesn't seem like they want to break the line.
This makes it mysterious since clearly CUDA is an advantage, but higher VRAM lower cost cards with decent open library support would be compelling.
Plenty of people use eg 2, 4 or 6 3090s to run large models at acceptable speeds.
Higher VRAM at decent (much faster than DDR5) speeds will make cards better for AI.
Intel and even AMD can’t compete or aren’t bothering. I guess we’ll see how the glued 48GB B60 will do, but that’s a still relatively slow GPU regardless of memory. Might be quite competitive with Macs, though.
Are there any performance bottlenecks with using 2 cards instead of a single card? I don't think any one the consumer Nvidia cards use NVlink anymore, or at least they haven't for a while now.
People actually use loaded out M-series macs for some forms of AI training. So, total memory does seem to matter in certain cases.
- less people care about VRAM than HN commenters give impression of
- VRAM is expensive and wouldn't make such cards profitable at the HN desired price points
Businesswise? Because Intel management are morons. And because AMD, like Nvidia, don't want to cannibalize their high end.
Technically? "Double the RAM" is the most straightforward (that doesn't make it easy, necessarily ...) way to differentiate as it means that training sets you couldn't run yesterday because it wouldn't fit on the card can now be run today. It also takes a direct shot at how Nvidia is doing market segmentation with RAM sizes.
Note that "double the RAM" is necessary but not sufficient.
You need to get people to port all the software to your cards to make them useful. To do that, you need to have something compelling about the card. These Intel cards have nothing compelling about them.
Intel could also make these cards compelling by cutting the price in half or dropping two dozen of these cards on every single AI department in the US for free. Suddenly, every single grad student in AI will know everything about your cards.
The problem is that Intel institutionally sees zero value in software and is incapable of making the moves they need to compete in this market. Since software isn't worth anything to Intel, there is no way to justify any business action isn't just "sell (kinda shitty) chip".