Search results

Why would it have anything to do with knowing anything, obviously having a personal assistant has a lot of value and Microsoft knowing stuff about an hashtag anonyme user that represent your session seem to have virtually zero cost (and a lot of it can be deduced from just your general internet...

GPU did not follow all the inflation trend it had his own explosion and they are now much cheaper now than 3 years ago. There new 4080 super around for $950-1030, maybe price will have known a local peak from the 6600xt at $380- 6600 at $330 until the 4060Ti 16GB at $500 was announced. Maybe...

Not so sure how different windows+c, typing a search will be versus launching edge and searching in bing in that regard. Obviously if you put the always look at my screen feature OpenAI offer on Mac or Windows on arm machine that will be coming to x86 with NPU we can imagine, but that more for...

Could depend on how much they can monetize the integrated AI search, which could be the focus instead of "in OS" ads outside the ads for their own product which for sure they will continue to push hard but those seem natural to the users like add-ons, they could be significantly better (because...

ML Server farm, surface or mobile Xbox ? That would make sense with the intel comments of AMD doing so well with x86 being surprised AMD would do this, but if it is a custom order with already a client....

Often have bigger power envelope ?, not sure if SIMM socket system are faster that the big GPU too, all being equal how could SoC memory not be faster exactly (being that it will be closer) ?

Video game benchmark for example can be real benchmark and be quite low stress on the CPU (like most game will tend to be for high core counts cpu), benchmark are not necessarily meant to be stress test (stress test seem better to stress test a system than using benchmark, those would tend to be...

They all can do change existing code. There is many way to feed piece of code to them, if you some made only for that purpose like copilot in visual code/studio/jetbrains/NeoVim, they will pass to the current context of the AI the current tab of the code you are working on and a little bit more...

That seem a bit strange, why reinstall the program that does not need to be installed in the first place ? Why is the CPU benchmark weak and why the GPU benchmark not worthwhile, i would have thought vray to be more interesting for GPUs than CPUs, asking because Puget use it quite a lot and...

It can depends on a lot of stuff. Human reading a lot of code and coming with codes has no legal ramification (sames goes for music, books, tv, etc...), normally inferred from trained material would be quite similar but they could come up with new rules. Some models can regurgitate 1:1 part of...

Could be wrong, but that seem a broad of a statement, will it not be just for those specifics APUs (and just for the new AI npu feature on those) ? While RDNA 3.5 gpu drivers will work fine on windows 10 for still a good while ? Windows 10s is still quite popular, but on new Laptop... ? It's...

Only if those step were made: "Check with the author(s) of the code, make sure that they were the sole author of the code and verify with them that they did not copy any other code. And inference generated code is not banned either according to the very excerpt, only in the thread title.

The fact that it will be a fork called FreeLLama open-source project with winamp (and importantly the mobile one) staying fully close, make it sound more possible. Stuff can be quite public even if no one can make code change in the official trunk (think sqlite, open domain not open...

Are they not like 80% of the market, what important enterprise did they ever ban ? I think you are trying to fit stuff a narrative, the idea that Internet explorer was ever driving too much for the next months profits. The always online was as much a long term vision than a short one, monopoly...

I had more people using their UHD disk file than streaming download

OLED 120hz 3K monitors with the Asus name will tend to be at least somewhat expensive, if that what the base model come with.... sound like spending a lot of money for a Intel Adreno GPU device, not that I know what that is.

I really do not how how it works but the last time. RDNA 2 launched (November 18, 2020) almost at the same time has the PS5 (November 12, 2020) RDNA 4 will launch probably around the same time has the ps5 pro or what they will name it. Would not surprise me if they go with AMD that PS6-XboxX2...

Should be way more in time than the next generation ? PS6 next Xbox could be fall 2028, fall 2027 would be in line with previous generation lifetime of 7 years. RDNA5 I would be quite surprised if it is later than 2026 (would guess fall 2025), imagine if AMD top end is the 7900xtx in spring of...

Maybe but their nuclear fission project is not short term profit either. https://www.cnbc.com/2023/09/25/microsoft-is-hiring-a-nuclear-energy-expert-to-help-power-data-centers.html https://jobs.careers.microsoft.com/global/en/job/1627555/Principal-Program-Manager-Nuclear-Technology Cancer...

Depends where in the stack the 4060 launched with a lower MSRP ($300) than the 3060 ($329), sometime it is minimal change while the cost stayed about the same, instead of minimal change + price creep like the 4060ti-4070.

Sure if you chase native 4k it can be quite something (always has). But at reasonable resolution and settings, as long cheap game consoles are popular and game studio make games that can run on them it is hard to imagine this ever be the case. Special for PC edition niche title yes, maybe...

There was window when they did not feel it was the case 9700 pro, 1080TI, 8800 GTX, 680 Take the 7800x3d arguably the best CPU to play game on that exist, better than the $10,000 cpu, they will say it is a reasonable price (even if it price by mm2 is way more than a 4090, the 4090 has like 8.5x...

HardForum not CompromiseForum

True enough outside chasing very high setting at 4k, say the high-1440p classic affair, how many worth to play game would not run perfectly fine on a 6800xt/7800xt ? Maybe Hellblade 2 will be one will see (if 30 fps do not play well), end of the days all the new big games has to run on 2020...

Even without a portal they could still win it (if they power OpenAI and OpenAI win the iPhone battle), sinking resource of what look a lost cause since say 2014 just for a possible 12 years later, they lost that war because they were very badly run (in part by the US government) when it was time...

aaaaaaaa got it...

Did those board had a bios flashback button and did you try to install the updated bios with it first ? https://www.asrock.com/support/QA/FlashbackSOP.pdf You do not need to have an installed CPU normally, let alone the current bios having any support for it.

The Amazon one during the primes days can get really cheap for the hardware, easy to "sideload" app on them. If it need for Windows on ARM to be good enough for it to work, it is a big risk, Qualcoom seem to have a list of issues going on. You can execute really well and still loose, Windows...

I read to me the: Even without it some boards let you update at BIOS boot without compatible CPU as CPU is just need for basic use (needs to be pin and electronically compatible). Flashback works without even need for a CPU in it as you don't need to switch on the system. Is exactly what I am...

Betting on fusion, by investing in it, doing R&D yourself, is all a bit of the same, those giants company tend to be the only looking very far ahead (most of other companies just try to survive). https://query.prod.cms.rt.microsoft.com/cms/api/am/binary/RW1fApf Look at the crazy stuff google...

It is somewhat common for motherboard to have their own memory-compute strong enough to do a BIOS update without a CPU or ram in the board just need power, almost all AM5 motherboard support it. You can look for bios flashback support or other naming with bios flash in them

The company doing nuclear fusion work, quantum computer and openAI is all about short term profit an underestimating the future is a really strange position... what's often less talked about is how much further ahead Microsoft would be if Satya Nadella They would be 4 billions company if they...

If they changed the precision of the AI model used for the simulation being run, they wrote it so small below the low res image that I am not even able to read it

Why say it is undocumented: https://opensource.apple.com/source/clang/clang-318.0.45/src/test/MC/ARM/arm-memory-instructions.s.auto.html Clang seem to support them and yhou do not need special library that take advantage of, something like that: char a,b,c; c = a+b; if(c<0) { } could use it...

Why would they ever remove said instruction, it seem to be a very small set of extra one. Even if we do not use any apple library (say a pure STL C project, nothing custom to MacOS), the compilers that target Apple Silicon could use those extra instruction, no ?

Chatgpt cannot run on your personnal hardware, they have over 200 billions weights (maybe in the trillion now) models running on supercomputer. The app are only on apple at first I think (and will use Internet to talk with a server), not sure if they are available to everyone yet. On your...

But why would they mind even at all if their binary run on other platform ? It is not like it is uncommon (that why people usually care about open source after all, how often you better compile specifically for your target...)

I did read like if it was about them adding stuff than the other would not be able to run anymore so people must buy their hardware to run their software, not for them to remove from themselve the ability to run application library. I was wrong rereading it, I guess it would change for people...

Would it be more if they get the same cooling, are they to the point that it does not even matter anymore ?

Didn't apple do this ? Not big but a short list of new custom instructions https://opensource.apple.com/source/clang/clang-318.0.45/src/test/MC/ARM/arm-memory-instructions.s.auto.html https://developer.apple.com/documentation/xcode/writing-arm64-code-for-apple-platforms Apple platforms diverge...