That's because the rival's prices are the most expensive in the world, with an RTX 5080 card going for ~3000$ on both Amazon.co.jp and Rakuten.co.jp
The prices are really out of control and common sense in that country...
As opposed to often ~$2000 in the US (I see there is current stock for ~$1600+). I've got to believe it's at least somewhat proportional in many places.
By somewhat proportional, I mean the nVIDIA cards are priced ridiculously and add almost zero incentive for the intended market (RT @ 1080p native/1440 up-scaling/4k up-scaling with FG).
If $600, even $1000 is beyond the scope of it's relative capability. The people of Japan can apparently do value and when good-enough (and not) calculations, which is even stated in the article wrt MH.
I understand that price is not generally $600 right now, but the other is also not $1000. The 5080 is double (and apparently triple), 9070xt is still not *really* comparatively nuts relative to
MSRP price/perf of 5080.
Hence why they're selling EVERYWHERE, even at those higher prices. Again...until they don't. But I truly do believe these cards are aimed at are tried-and-true AMD market of ~400-450/500-550, eventually.
Once they can meet demand, and who knows how long that will be, wafer allocation elsewhere or not. They've got a bit before they need to worry about any competition (from say a new console and/or 3nm parts).
That's when market-share should truly increase; by bringing that tier of performance to a new market and to more people unwilling to spend, say over ~$500, on a GPU, which there are many.
I like the conversation about aiming for 70% market.
To me that signifies AIBs want not only similar performance at lower pricing (as N48 is to AD103/B203), but better performance at similar/lower pricing.
And, FWIW, I do expect UDNA (if all goes to plan and both companies aim as they typically do) to be exactly that.
If it captures the actual sales, IDK, as the marketing distortion (re: certain features/scenarios of games optimized toward their arch [which are few in reality but heavily used by the press]) of nVIDIA is peak.
Are people waking up? They might me. They might not be. Will it happen generally; outside of places like Japan (which I often see as trendsetters)? Time will tell.
Doubt it. nVidia and their partners, make loads more profit of fewer cards sold, than AMD for their entire stock. And at the same time, nVidia, sells eight times more, with their completely slumberous Consumer/Gaming GPU division.
So MSI probably gains from the nVidia/intel exclusivity more, than wasting resourses on AMD-powered gimmics. At least this is more honest, because earlier MSI was just showing the disgrace for AMD products. Much like other dual-camp AIB partners.
Asus used to sh*te on AMD cards, just utilizing the existing coolers, designed for nVidia. The same goes for AMD motherborads, which, much like the graphic cards, were a tier below intel/nVidia counterparts. Gigabyte is not much different.
At this point, they all may abandon producing any AMD SKUs whatsoever, and still stay in profit.
A post by somebody that clearly does not understand the customers of each company. You should not see things as 'a tier below' but rather those customers buying 'what is required' and nothing frivolous.
It has literally
always been this way for AMD/ATi vs Intel/nVIDIA, but in different respects. Be it a 250w+ 14900k vs a cheaper 65-120w 7800X3D or a 1080p60 RT card vs a way more-expensive 1080p66 card.
The AMD choices literally similar uses for much lower cost, sometimes with real advantages (like a 7900xtx vs 5080 in 4k raster), and every piece of the ecosystem is similarly-targeted toward people w/ that ideology.
This has been true for
literal decades and for different reasons. Since the Athlon XP and (9700pro, but really) RV670. Bulldozer and Fiji/Vega excluded, as I don't see them as symbolic of the overall brands.
The thing MSI is going to find out is that nVIDIA adjusts *their* prices each subsequent generation to capture any gains the AIBs may have made in the former, while limiting supply to keep best unit/margin ratio.
I think AMD stays pretty consistant in their market aims and general costs. ~$250/500 (perhaps currently $300-600 to capitalize on market situation), with each satisfying a market (1080p/1440p).
A halo that competes with whatever nVIDIA has one chip down. Similar market aims as one tier up from nVIDIA, but perhaps some concessions to keep those prices (ala RT/ML until now, maybe PT/bw now).
But still satisfies the market of 4k while remaining in that generally-accepted pricing structure. What happens when nVIDIA can't sell their parts at riduclously high comparative margins for a company like MSI?
One way or another, nVIDIA will find a way to make the same or more money, partners be damned. This has been proven over and over again. If relative margins falture, the AIBs will be the ones taking the hit.
All-the-while AIBs are clamoring for allocation (for those margins to keep them afloat), where-as nVIDIA will constrain it to keep ASP high (as that benefits them most, especially when silicon can be used elsewhere).
The problem is AMD didn't have GPUs that people (I would say understood correctly, but regardless) wanted to buy. Now they do, and will likely continue going forward given the leaps to catch up in RT/FSR.
While path-tracing might not truly be on most peoples' radar for quite a while (perhaps the generation following the next one, if not the next major node shrink after that). As such, AMD's position will likely improve.
Now they simply need to hit the 1080p, 1440p, and 4k RT markets (and with the former two up-scaling to the later in others). And they most certainly will. Likely for similar prices as always.
nVIDIA will try to find another gimmick outside general use-cases for margin, which this time is likely path tracing (unless AMD too adds more bandwidth/cache to their next arch and they are similar).
If they'll succeed, I don't know. Many don't understand that when those features debut they are designed to be outdated consistantly until a standard emerges (as we see with RT/older cards using newer DLSS).
And this how both companies sell cards; by 'progress' in a new field that consistantly is outdated each gen until standardized, or selling that standard (and nothing more) as affordable (and to best TAM) as possible.
For a company like MSI, as always, I would argue sales volume is important above all-else. I feel they will/would get it from AMD (especially in the future), where-as from nVIDIA this is much more doubtful.
