Intel has announced that the chipmaker has discontinued the Arc A770 Limited Edition gaming graphics card.
Intel Discontinues Arc A770 Limited Edition GPU : Read more
Intel Discontinues Arc A770 Limited Edition GPU : Read more
AMD had CPU chips as their core business.Wow. AMD didn't give up when their first generation of x86 chips weren't up to par.
Unless I missed some bit of news it doesn't sound like Intel is giving up on the dGPU market. The article statesWow. AMD didn't give up when their first generation of x86 chips weren't up to par.
Instead, the chipmaker can now prioritize silicon production for its AIB partners
I had their first gen GPU the i740, and while it was faster than their first IGP (the μPD7220 licensed from NEC), decent was not how I would describe it, not even for 1998. They thought AGP was so fast that there wasn't any need for much local memory.Actually, for first generation GPUs, Intel's were quite decent!!
I hope the "refresh" still happens. Maybe that's why they're trying to whittle down their inventory.The A750's price bouncing back from $200 to $250 was a pretty solid hint that Intel may have been tightening supply.
Wonder how long supplies are going to last. Then again, I don't imagine demand being particularly high.
The A7xx performs anywhere between roughly as expected from the specs to only a laughable fraction, which means the hardware is able to perform as-is when the software planets align well enough.Even with the latest drivers, it's still performing well below where its specs would suggest. So, I think it's plausible there are some hardware bugs holding it back.
Many of us are still on the fence with Intel graphics cards.
This is their third foray into the D-GPU market.
They abandoned their first attempt shortly after launch. about a year or so, then abandoned the second attempt after releasing it as pro only.
Will Intel abandon their customers again? for the third time?
Companies are always releasing better products. Maybe it's a redesign, but sometimes just a refresh. Early adopters get to enjoy newer tech for longer than people at the trailing edge, so it's not as though they got nothing for being early. Also, when the refresh launches, it'll cost more than the original revision cards. Finally, the original Alchemist launched against RDNA2 and Ampere, whereas the Alchemist Refresh needs updates to compete against RDNA3 and Ada.If Intel launches v2 Alchemist silicon that performs far more predictably than the v1 stuff, people stuck with v1 hardware may get pretty cheezed about it possibly to the point of starting a class action for their known-bad-wontfix hardware.
If Intel releases an Alchemist refresh that performs significantly better than the original almost exclusively due to fixing hardware bugs, then it means the original hardware was fundamentally broken and first-gen buyers may have a case for being sold defective goods.Companies are always releasing better products. Maybe it's a redesign, but sometimes just a refresh. Early adopters get to enjoy newer tech for longer than people at the trailing edge, so it's not as though they got nothing for being early.
If the "bugs" only impacted performance, then it would be hard to litigate, because Intel never guaranteed any performance levels of the original product and would simply characterize the fixes as "improvements".If Intel releases an Alchemist refresh that performs significantly better than the original almost exclusively due to fixing hardware bugs, then it means the original hardware was fundamentally broken and first-gen buyers may have a case for being sold defective goods.
Intel has stopped selling their reference card, LE stands for limited edition after all. They have made it clear that they are continuing to sell to AiB partners.Not surprised. Intel hasn't pushed their cards on the market, I don't recall seeing commercials, the driver blunder at launch, etc..
This isn't the right move. This tells customers "we've give up" and User confidence will be zero in the future.
What risk? It isn't a computing ecosystem.That's fine but if first party (Intel) puts no effort into their product, who will want to buy it from a 3rd party/AiB?
Its like buying anything from Google and having them discontinue it a year later. You are making a risk buying into an ecosystem (Intel d-GPU) that is barely established. I think it would be wise for them to at least replace the LE with something slightly comperable with LTS
I also dabbled with Linux, pre-1.0. ChromeOS isn't ready for prime time?I didn't include my various leaps into linux, but my 1st version was a pre-1.0 kernel. It will never be ready for prime time.
The risk in the meantime I are drivers. They failed epically with them at launch and took forever to fix them.What risk? It isn't a computing ecosystem.
A computing ecosystem is os based. Let me run down all of the computer ecosystems I have moved through:
DrDos/Geos (1988 - 1990)
DrDos/Windows 3.0 (1990 - 1992)
OS/2 v. 2 - 4.52 (1992 - 2001)
OSX 10.0 - 10.10 (2001 - 2019)
Windows 10. (2019 - present)
I didn't include my various leaps into linux, but my 1st version was a pre-1.0 kernel. It will never be ready for prime time.
We are talking about a video card running on Windows. I have moved from back and forth between AMD to Nvidia for years, and now I am trying Intel.
I spent $349 for a 16Gb video card that performs (currently) between an RTX 3060 & an RTX 3060ti. I'll be getting driver improvements for at least another year. I have a high refresh rate 1440p panel, and I don't see myself moving to 4k anytime soon.
If battlemage never releases to desktop, what have I actually lost? Nothing.
A few years down the road, it will be time to retire the card, and I'll see what I need, what is available, and how much it costs.