If the first gen’s prices go lower than they already are, I’ll buy the hell out of an A770 for video transcoding in AV1.
Intel cards are not the best performance for gaming, but they’re getting better in a lot of respects, and they are legitimately the only manufacturer targeting the midrange and low budget segment. I hope the second round of cards are significantly better while staying within the same price range. I also hope the standalone cards help them tune the Core series better. The MSI Claw is a fucking travesty.
If your only intention is to use the card for encoding, I recommend picking up an A380 instead. The A770 is a surprisingly performant gaming card for newer titles, but all of the available ARC cards have the same encoder.
Since the A380 is typically single slot, and fits within the 75W spec, you don’t even need an extra power cable for it if you wanted it as a secondary card too.
The problem with Arc cards is that they are a great budget option but don’t work on budget computers. I was planning on pairing an arc 380 with a i7 6700k for a little budget gaming machine but the card doesn’t work without ReBar. So who are these cards for? Who is pairing a $100-200 GPU with a $200 cpu?
I don’t know how they plan to compete in the budget market if people are better off buying an 8 year old GTX 1060 then they are buying a new arc card.
Intel seemed to fall behind kinda hard w/ regards to CPU/Motherboard features until much later on. Supposing you aren’t working with parts you already have, everything from 1st generation Ryzen onwards would have rebar support. They can be had very cheaply too, and work on any AM4 board.
You may also find a BIOS update allows some older chipsets to support rebar. It’s a tad flaky depending on the manufacturer though.
There is a hack to get ReBar on those older machines. I have a 6700k as well and it’s running a 1050ti, but this utility I heard about allows any modern system running a UEFI to have ReBar.
Alchemist is barely more than a proof of concept. Until they got DirectX to Vulkan translation really going, the performance in a lot of titles was kind of a joke.
A gen 1 product only ever makes sense in hyper-specific use-cases, but a lot of the pieces for something really neat and much more widely useful, are there.
Alchemist cards weren’t gamechangers. But they’ve proved that with time and some more RnD, ARC can be.
They didn’t WANT to make budget cards. They wanted mid/high end cards but fell short. But it’s better to sell at a lower price then to completely eat the cost for hardware they already made.
Add comment