They already said the hardware engineers have moved on to Celestial

APO is the key. There are already number of titles and I think they are looking to have more before launch. Also to have APO the MB's available and BIOS updates are still in short availability. If you don't have MB's that can handle it it is too soon. We have a confirmation now that 12th and 13th gen CPU's are included for APO. Initially it was only 14th gen CPU. But as I said the BIOS is missing from most of the MB's to give the APO. APO can give 10-50% boost so not a small gain. By waiting you are forcing the MB manufacturers to work on the BIOS.

🤣

GPU Consumers: Thank you! Intel has freed us! Intel: Oh I wouldn't say freed... more like... UNDER NEW MANAGEMENT!

done in November 2023 with A770 and its nice :)

i buy intel arc gpu all the time so im a suporter 🤣

And anti consumer VRAM amounts.

@@christophermullins7163 i brought new 3070 ti for 300 usd, yet i feel 8gb vram is such a turn off

and anti consumer and anti competitive practices.

also you might not yet know how true this is/can be if intell would publish battlemage soon with the extra thing which which made you wait now, I hope they finnaly made up their mind and finalized the choice(expect they did, if they dind't by now then they would be waiting for arrowlake, but I sure hope that they will do battlemage now and let celestial wait for arrowlake instead(since celestial is after battlemage anyway)). if they do then unless nvidia already is planning to do something similar then nvidia will have no more place left on the market, perhaps only as some high power usage super expensive high performance gaming gpu, but in terms of AI, normal gaming and power efficiency nvidias current lineup would look like litteral poop if they continue doing that thing in battlemage already, while many might not notice or think about it to much now, it actually might be a much bigger change and advantage than the ones the people and leaks now know and talk about, ofcource if they continue with it instead of dropping it and chosing to instead only put it in arrowlake and such since it requires some optimizations to move it out of the cpu/apu and make it work as well on a dedcated gpu which has to communicate through a pcie lane which is relatively slow or rely on the Vram which often is used for other things, well perhaps that other thing might fix that as well since while I have never really seen hardware acceleration vram compression on a gpu, it does make a lot of sense since vram needs high bandwith and this can actually both increase technical capacity as well as bandwith. but for the other thing I am mostly interesting about, you need to look at arrowlake, lunarlake, meteorlake, phoenix. on first actual look you might notice a pattern but that won't say much, only when you also look into certain FOSS softwares, and open standards which either where set up by intel or certain interesting contributions made to projects by intel as well as what they heavily focus upon, perhaps some last tips: italian wine and 1, that should easily allow you to see 2 speciffic open source projects which intel has set up and put insane amounts of work in recently. if you still don't get t then ask yourself the question how a simple low power APU from arrowlake can beat a rtx3070(like you have(depending on the version of rtx3070) in AI like stable diffusion as in similar, very close performance while using insane amounts less power, don't know/think those apu's will actually beat the spead of a good rtx3070, but I have seen people posting their it/s in stable diffusion from a rtx 3070 and between those there where actually quite some which where close to the apu range, still don't know the apus actual full performance only a minimum improvement and it's insane efficiency, the same goes for the rtx3070 since they seem to have a very wide difference in performance in stable diffusion, perhaps nvidia published multiple different gpu's under the same name or it is the difference between some oc or gaming cards and some budged cards. the APU however should greatly beat them in efficiency and still get close enough in performance that it can generate images in a few seconds(rather usable since online AI tools like bingai already let you wait longer ), while using almost no power to do so. rtx 3060 would almost certainly be beaten in all aspects in the field of AI(AI also affects games like AI game enhancements, AI game optimization(which wasn't yet done much due to AI being heavy to run), AI superresolution or superdetail, worlds which become dynamic and alive due to ai, etc. and do not forget that if you can run AI far more efficient the AI in games won't affect your free compute power as much) but that is a low power apu, so if they end up indeed using the same tech already in the battlemage GPU then we might assume there is a rather decent chance it will be far more powerfull even though I don't know if they will still hold back with battlemage, or if they seriously go in right now, the holding back somewhat already is fun since even if they add 4 times as much of that special stuf in the gpu than in those APU's(and in those APU's it is actually only a very small part of the chip die, as well as rather low cost in the total price), so 4 times is rather doable, but at that point basically all nvidias current gpu's would lose any of their unfairly reached advantages for things like AI(finally). while if they actually go more deeply/seriously into it instead of rather conservative then that would litterally make all nvidia gpu's pretty much useless even their next gen(in things like AI), unless they happen to do the same thing in them. ofcource this is assuming intel can manage to get it into it in time and it will work well enough/is doable enough for now since this is kind of a rapid push.

No Nvidia own Intel at the moment.

yep might be announced then, I hope they will do so on intel vision since that is just over 1 week away. also makes sense for them to announce it then, since they already mention a focus on AI, meaning they will likely talk about Lunar lake, but they already did, so there is more, they'll focus on. I sure hope however they have already made up their mind surrounding the amount and level of NPU power they will add in their battlemage gpu's if they have by now already made the official choice to use it in battlemage already instead of next gen or only the cpu's/apu's or dedicated cards. I really hope they already made up their mind instead of waiting to see how well arrowlake and Lunarlake will do in order to decide it. if they make the tpu 4 times bigger than arrowlake then technically seen if it can use the systemram it should get better performance than a rtx 4060ti 16gb in stable diffusion while also barely using any power(and not even using the gpu itself, only the NPU which is a small low power chip) in comparison. but I am wondering if they will go with something similar to what they put in arrowlake, something 2 to 4 times faster, or something actually big causing it to finally purge nvidia from the AI game since then they hold not only far greater efficiency but also far greater performance. or if they decide that it will only be in the cpus/apus, even though I personally hope they will either go for around 4 times arrowlake, or even bigger to seriously beat the world market. also if you are a gamer, and don't run things like stable diffusion, then see it like this, such NPU's use almost no power and are very small in size and also very cheap to make, but they can run AI tasks insane amounts faster and more efficient than even the AI optimized gpu cores, this mean things like AI upscaling could also be many times better and faster and not actually affecting your gpu or power usage as much. next to that even the current old generation integrated NPU's which are of the same power as the NPU's you sometimes find in some raspberry pi or arduino competors are capable of running not only stable diffusion and such but also entire modern language models locally at low power usage and pretty much real time. so things like AI npc's which nvidia wanted to push, can be made to run fully locally with cheap hardware(since the npu's tend to only be a very small part of the price when looking at the intel and amd apus which already support it) but technically seen you could run such things much faster and more efficient on such new intel harware without it really affecting your gpu power, while the way nvidia does it does affect rendering and especially raytracing power.

yes it is they really needed to get it out ahead of RTX 5000 /radeon 8000 series of gpus, because all data on it looks like it will compete with rtx 4000/RX7000 capability wise ... bassically putting BM a gen behind the up coming stuff from nivida and AMD. but then again what intel doesn't need is another arc launch, with drivers all over the place like a mad man living in the forest. so maybe it's good they are taking this long and hopefully theya re geting the drivers rock solid before launching it this time.

​Intel : AI AI AI!! 🤓​ thats what also nvidia and amd is about so wait for them to sxck your blood for AI and then you'll get what is promised for you. Whic you dont know when it will happen it can take anytime at this point.@@DenverStarkey

@@DenverStarkeyBattlemage will have much lower prices than RTX 4000, while having significantly better RT performance than RX 7000. Whether Battlemage will be competitive or not depends on the reliability of their drivers and how good RDNA 4 turns out to be. I think we can both agree that Blackwell’s performance advantages relative to Battlemage will likely be offset by its expensive pricing.

i disagree with this as arc released after first the 40 series gpus and a few weeks before the first rx 7000 series

@@Vorexia true .. but still it puts intel behind. intels one saving gracve with taking "too long" would be that they get their drivers nialed down good this time. they manage that , the "late to the party" and it's price will certainly make intel more compelling.

Why would you possibly think they would be more expensive than the other two?

@@NahBNah Judging by their current line up don't think so...their CPUs yeah but GPUs I think it'll be cheaper, coz competition.

Watch the previous video of red gaming tech u will know why

No thanks he is clueless. The fact you are unable to explain tells it all. If you have a real point tell it or shut up.

@@K543 👍

Even AMD users get a lot of hate so I cant imagine the hate you get lol I guess the price was really good?

agree my arc 770 16gb founder edition is awesome used it mainly since launch and it was 300 bucks

A 32gb version would be wild

But the 4090 is 24gb 😭

They might be just adding some final details to the cards because the stated once that they finished with the software side and were just working on the hardware side

battlemage is going to beat the 5090

@@badpuppy3 lol no. It will barely beat the 4070 super.

Or they might just release crappy drivers and then everytime they release new ones they can claim their GPUs are gaining 100%+ more performance...

​@@badpuppy3🤣🤣🤣 Yeah, no... They will need to overcome impossible generational leaps to do that. I've been running the A770 for about a year and... It's having trouble competing with the 3060Ti... We'll be lucky if BM gets anywhere near 4070Ti/7900xt.

Gonna call the customer service department?

@@christophermullins7163 few emails to engineering should do the trick😀

Whether the 4090 will be mid-range or not depends on how mid-range Blackwell GPUs perform. Even if the 5090 turns out to be a monster, that doesn’t guarantee that the 5070 won’t suck.

Lol if the 4090 is a mid range card I will sell my pc. I highly doubt it. It will still considered high end. You need 70 class card to be better. Issue is that won't happen.

Nah, the 4090 will be just high end from enthusiast class. RTX 4090 will become midrange when the RTX 6090 launches.

@@raven80wolfx2 RTX 5080 is gonna be the same or slightly faster than a 4090 so 4090 will still count as high end.

Lol amd is focusing on mid or low end. Intel is too. Nvidia will be high end and idiots will still buy nvidia. They don't want competition. They want the best and they pay more for it. I like to say that people waiting for new card. Issue is 3060 is still the most bought card. Even if right now we have better and cheaper cards. It's still what most people are buying. Amd has better price to performance on all fronts. Intel can beat a 3060. Just depends on the game. The reason why people buy nvidia because it's all the know. They won't have as many driver crash due to Intel being new to the game and needing rebar support as a must have. I switch to amd finally. If it wasn't for them I wouldn't have been gaming during the pandemic. Fsr was a life safer for me. Depressing world we live in when people are just brand loyal. Even amd cpus still haven't reached top percentage yet. Intel has too many deals with companies to not be number one.

​Im build pc on my own for the first time and i would like to know which gpu should i go with. Rx 6600, arc a580 or something else that ain't cost like 100€more

@@fubkegt4964 Forget the arc580. Too many bugs still. Go with the RX 6600.

Intel said 4070 so .....

An rtx 3070 is only about 40% faster than the arc 770. 2.5 times the performance should be around a 7900 xt or an rtx 4070 ti

@@michaels9917 we will see, certainly hope intel catches up

Why dont u see Intel selling much?

@@gamingmax9455 we are getting to close to the launch of the 50 and 8000 series. So unless Intel has a great product and a great price most will just wait till the 50 and 8000 series release.

its too good to be true 😅

Watch the gpu plate fall off

@@triphecta lol. Also when they are ready to compete in the high-end market in the Celestial generation, they should have a C919.

Negative, their release target was always late Q2.

Lower them all by $50-100 and that would be the sweet spot

@@Someone-yt8gn Nah wouldn't be profitable for them that's absolutely stupid

You’re dreaming

he also said that alchemist is not going to make it, but im typing this on a computer with a750

That fool is wrong all the time and deletes his videos instead of just owning it

He's been wrong _every_ step of the way...

that's what people used to say about AMD and CPUs

depends on price, amd isnt going highend next gen so intel can compete in mid

Me 🤑

No one cares about Intel GPUs. Everyone cares that there is more competition to keep price to performance progressing on your favorite AMD or Nvidia GPUs.

Anyone with common sense cares.

I care because im a value shopper