• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Intel Arc "Alchemist" Desktop Graphics Cards Launch Pushed to Late-June/July

well then im definetly waiting for rnda3/ada
 
If the drivers are stable (and this is where their iGPU drivers excel - they are stable), even if not really very performant at first it will be enough.

I think it is reasonable to warn those who will buy into Arc first gen that they should be very mindful that they likely will not have NVIDIA grade drivers, but imo it will improve significantly over time. For a geek like me, that just screams fun :toast:
I totally get you. :) Part of me wants to buy an Arc GPU even if it's bad, just to play with it and satisfy my curiosity. :D
 
That didn't work out too well for Matrox and their Parhelia. At one point (early 2000's?) Matrox was a player in gaming GPU's.

Except Parhelia launched late, in an utterly broken state, AND way overpriced.
 
I totally get you. :) Part of me wants to buy an Arc GPU even if it's bad, just to play with it and satisfy my curiosity. :D

These will be the first genuinely new things in graphics for quite some time... RDNA 3 and Ada Lovelace might perform but they will both be priced to match and just iterating on the existing stuff. I think they will be exciting, with the kind of demand that we have out of PCs and hardware in general I would even agree if someone called it unwise to main one, but they will be wonderful playthings... besides, not like my 3090 is going anywhere or becoming unusably obsolete any time soon eh :D

But Dr. Dro fun doesn't necessarily translate into sales -- otherwise the hula hoop would still be a thing.

Price makes or breaks things (of course usability also plays to an extent), but I am interested in Arc from the point of view of someone who likes to understand how tech works. It will deliver that :)

Larrabee 2.0 ?`??​


I honestly do not think it will be Larrabee all over again, and these things are not Raja Koduri's fault. That copium of blaming him for every misfortune related to the AMD products he has worked on in the past decade has grown quite old and I really wish the redditors that perpetuate that myth reconsidered it
 
These will be the first genuinely new things in graphics for quite some time... RDNA 3 and Ada Lovelace might perform but they will both be priced to match and just iterating on the existing stuff. I think they will be exciting, with the kind of demand that we have out of PCs and hardware in general I would even agree if someone called it unwise to main one, but they will be wonderful playthings... besides, not like my 3090 is going anywhere or becoming unusably obsolete any time soon eh :D
I don't intend to replace my 2070, either. I just want an Ark as a secondary plaything. :D If it really ends up being a failure like some people "predict", then it may as well become a collector's item in a couple years, just like the aforementioned Matrox Parhelia did. :)
 
Who's to say Arc won't have a similar launch?

It's a possibility. But the original stipulation in the comment you replied to was, "if the price is right." Parhelia's price was definitely NOT right, and I have a bit more faith in Intel to at least offer something in their lineup that people could recommend over their contemporaries.
 
It will never release. Driver issues and a comparison to other vendor options mean this is dead in the water.
 
I just assumed it was never coming so,.....
 
It doesn't matter how obsolete they are at launch if their price is right. It's only that Intel won't be able to sell them for a price they'd potentially want to sell them for.
Intel would still sell them where nvidia is, remember this is intel we are dealing with, they don't gaf about pricing, if its good enough for the goose its good enough for the gander.

Who's to say Arc won't have a similar launch?
Matrox got snuffed out just like PowerVR.
 
These will be the first genuinely new things in graphics for quite some time... RDNA 3 and Ada Lovelace might perform but they will both be priced to match and just iterating on the existing stuff. I think they will be exciting, with the kind of demand that we have out of PCs and hardware in general I would even agree if someone called it unwise to main one, but they will be wonderful playthings... besides, not like my 3090 is going anywhere or becoming unusably obsolete any time soon eh :D



Price makes or breaks things (of course usability also plays to an extent), but I am interested in Arc from the point of view of someone who likes to understand how tech works. It will deliver that :)



I honestly do not think it will be Larrabee all over again, and these things are not Raja Koduri's fault. That copium of blaming him for every misfortune related to the AMD products he has worked on in the past decade has grown quite old and I really wish the redditors that perpetuate that myth reconsidered it
Why wouldn't the problems be his fault? He's heading up the project.
 
Intel would still sell them where nvidia is, remember this is intel we are dealing with, they don't gaf about pricing, if its good enough for the goose its good enough for the gander.

Matrox got snuffed out just like PowerVR.

If Intel doesn't GAF about pricing they may find themselves sitting on product for a long, long time considering the stiff competition they'll be facing in that arena.

Surprisingly, Matrox still exists:
https://www.matrox.com/en
 
One easy assumption to exclude is that they are delaying in order to polish their drivers to achieve acceptable performance in key titles in order to get favourable reviews and price them up accordingly.
What good this strategy will do them if they are forced after one Q to revise their pricing down, they just would have lost precious momentum to gain positive mindshare by launching with aggressive pricing and trying to force the competition to follow.
Probably the state of software/drivers right now is not were it should have been, not (only) from a performance standpoint but more importantly regarding stability, bugs, etc and launching sooner than later could risk permanently damaging the multi-year arc affort, that's my guess, it shouldn't be availability related issue.
 
Why wouldn't the problems be his fault? He's heading up the project.

I haven't seen anyone blaming David Wang or Jensen for the state of the semicon industry right now, so why blame Raja? If they are having issues with building enough cards and delays are necessary, then so be it... Mobile Arc has begun to ship and is slowly appearing in many countries, so desktop cards are just a matter of time.
 
I mean, it's a bad plan marketing wise, but the big money is in the lower end volume side of the business, not on the big halo stuff. But yeah, they'll have a hard time setting their price, might end up as very limited availability on the desktop and dump most of the stock on laptops where they have much greater pricing power (because they can bundle them together with CPUs)
This is demonstrably wrong, as proven by nvidia constantly, the big halo chip market is far larger then credit is given, and makes the majority of the profit. They proved this with the 10 series, despite only the 1070 and up bei by released, they outsold the entire rx 400 line by nearly 4:1, and at over double the average price.
The low end volume market has been obliterated by APUs Colbert the last 10 years
 
I honestly do not think it will be Larrabee all over again, and these things are not Raja Koduri's fault. That copium of blaming him for every misfortune related to the AMD products he has worked on in the past decade has grown quite old and I really wish the redditors that perpetuate that myth reconsidered it
No something is out there, i just forgot to add the to my post :D
 
This is demonstrably wrong, as proven by nvidia constantly, the big halo chip market is far larger then credit is given, and makes the majority of the profit. They proved this with the 10 series, despite only the 1070 and up bei by released, they outsold the entire rx 400 line by nearly 4:1, and at over double the average price.
The low end volume market has been obliterated by APUs Colbert the last 10 years

Steam hw survey (far from a perfect source but good enough) disagrees, there are about as many 3090 and 3080ti combined as there are people running Apple Silicon (this one really blowed my mind lol). The higher end market has much wider margins but I doubt it can account for anywhere between 2x to 10x differences (again, steam is not the greatest source ever but looking at numbers like 1060 at 7.95% - 1080 1.43% 1080ti 0.84% or ampere 3060 2.40% and 1.23%)
 
These will do nothing to lower prices...
why ppl via the steam server hardware millions are using the 2016 `1060 still gpu and millions more are using weaker... what do you do for a lving in tech or supply/demand that more of somthing better would help prices that have been droping alot that last 4 months... s/d means everything its the reason for the massive and med cost of gas for the decades... Ukraine kartina covid 9/11 ect ect... are you that dense?
 
why ppl via the steam server hardware millions are using the 2016 `1060 still gpu and millions more are using weaker...
Because
1. Those are still okay for 1080p gaming,
2. There hasn't been any great invention in computer graphics in the last couple years that would necessitate something better,
3. The general trend where the gaming industry is heading is disappointing to say the least. Nobody wants to play the remake of the remake. People want something new, something original which is quite rare nowadays,
4. The Steam survey is done globally. Where the majority of people live (poor or developing countries), people can't afford a graphics card at "component shortage" prices.

My 2 cents. :ohwell:
 
Back
Top