Nvidia is also investing $5 billion into Intel
What just happened? Intel has received another massive investment from an unlikely source: Nvidia. Team Green is purchasing $5 billion in Intel common stock at $23.28 per share, part of a collaboration that will see the two companies jointly develop x86 system-on-chips – called Intel x86 RTX SoCs – that integrate Intel CPUs and Nvidia RTX GPU chiplets for a wide range of PCs. Intel will also be building custom x86 data center CPUs for Nvidia to integrate into its AI infrastructure platforms. […]
The article mentions nothing about nvidia giving up on WoA and I highly doubt they have. This seems like a great choice to give people what they want in the short term while they work on bringing WoA to reality in the long term.
x86 dominates consumer desktop/gaming space now…but i believe long term x86 is dead. The future is arm.
But they are hedging their bets (or perhaps being pressured by the US administration to partner with Intel).
I will strongly disagree with ARM being the future and x86 being dead in the long term (we’ve heard this for over a decade).
Beyond gaming and peripheral drivers (even though this is massive issue), there are multiple other problems with WoA that tend to be ignored:
Don’t get me wrong, I’ve been using ARM for 7 years in my Raspberry Pi DIY home server, but the above-mentioned issues are real. With ARM competing directly with its license clients, it’s more likey RISC-V is the more viable long-term option.
x86 dominates today. Its a smart move.
I don’t think legacy software will keep x86 around and winning forever. Things change and its not always easy to see it before it happens. I believe all the reasons you listed will keep x86 around for quite a while.
As I told another, I want RISC-V to win so badly
I would argue it is incorrect to categorize x86 software as “legacy software”. Outside of some Apple ecosystem products (which benefits not from ISA, but from vertical integration), all software/peripherals treat x86 (windows specifically) as the primary platform, with WoA being completely ignored in the vast majority of cases or treated as a low priority, tertiary platform.
Look at the ratio of WoA only games to Win x64 only games. Or WoA only hardware peripherals; they don’t exist (just like WoA only games don’t exist).
I’ve been tracking the release of Snapdragon X Elite pretty closely. The benchmarks on the CPU side are disappointing to say the least. Qualcomm also knowingly engaged in what is de facto fraud by showing unrealistic preliminary benchmarks using custom cooled laptops devices and running a “for benchmarks only” build of Linux (look at the state of Linux support on X Elite devices almost a year after release). GPU benchmarks are an even bigger joke. Even with alleged battery life benefits of Snapdragon X Elite WoA failed to materialize on a “apple to apples” comparison basis.
On a market share basis, they barely unable to hit 1% during the immediate full quarter following release, a bit of strange situation for something that supposed to be competing against a legacy platform.
And keep in mind, market share data for shipments/POS is typically on a gross basis, market research companies don’t account for returns and don’t make historical adjustments based on returns (this would be impossible for shipment into channel data). WoA devices were noted to have higher return rates. Qualcomm says it’s not true, but they also have the capability (they have the data and the legal authority) to publish specific numbers and they didn’t do it.
Note, I am not pro or anti anything. I care about lower prices for the same level of performance or features. I am pro intense competition where margin tend towards zero and companies simply can’t increase them as all money goes back into R&D and maintain lowest prices possible. A real free market, not the polemical version pitched by oligarchs.
If it was up to me, I would change IP laws for computer hardware to make all relevant copyright assets (drivers, firmware, EDA files, internal tooling critical to delivery) and patents last 5-10 years. Essentially anyone would have access to everything needed to replicate a 2080 Ti (or Zen3 or M1 or Snapdragon 855) at the cost of materials (plus 2-3 % percentage). Literally running the Nvidia driver with minor changes to update the branding and copyright text.
When I say arm will win in the long run, I don’t mean that WoA will have anything to do with it. WoA as it exists today will likely be abandoned.
I do think in the next decade Microsoft will focus on creating a full abstraction layer so it’s software can run on any HW. This would be a longterm play. At launch, Would it be feature parity with what exists today…no. but over time it will build up the codebade, drivers, etc. To over take the current windows ecosystem.
*RISC-V
I want RISC-V to win so badly.
I think it will since China has began to fund support for it pretty heavily.
It’s a bit funny how China has been heavily investing into open source, but I guess it makes sense - it’s tech they can use readily and save money on using while also disrupting the profits and influence of western corporations. It’s clearly for political reasons but hey at least the open source community wins
Has there been any disruption of profits and influence of western corps due to RISC-V? To my knowledge, not yet and its not a sure thing.
Im not entirely sure your point of view is a win for OS. When other nation states view it as a threat instead of an option it leaves the door open for all the sabotage, etc. that nation states to to each other.
Not yet, but something like Deepseek did disrupt some profits. If RISC-V keeps developing and lowering in costs it my may do the same, but as of now it’s still pretty pricey and experimental in use.
I’m just talking about the hardware here. Hardware is much more agnostic, and the costs going down + it being open means other countries can eventually manufacture their own at a lower start up cost and increased profits compared to say ARM.