Intel plans another big graphics driver improvement in February

Daniel Sims

Posts: 1,365   +43
Staff
Something to look forward to: Intel has had a hard time entering the dedicated PC graphics card arena. Its first GPU series struggled to reach the market when the company said it would, and initial impressions revealed severe flaws. However, Intel has done a lot to fix some of those deficiencies and remains determined to improve its dedicated graphics offerings.

Anonymous sources tell PC Games Hardware that Intel is preparing to release another significant driver update for its Arc graphics cards in February. Significant changes might also soon come for the company's graphics control software.

The update should drastically improve Arc Alchemist's performance for games using all graphics APIs. The patch follows a December update that improved the GPU line's massive disadvantage with DirectX 9 (DX9) titles.

One of Arc Alchemist's main weaknesses is that it doesn't natively support DX9 because Intel needed to focus its development efforts on DX12. Games started transitioning from DX9 to DX11 over a decade ago, but some of the most popular titles, like Counter-Strike: Global Offensive, still use DX9. Alchemist GPUs initially ran those games far worse than comparable cards, but the December drivers made up much of the difference.

The February drivers could repair the deficit further while also improving DX11, DX12, and Vulkan performance. The update will also introduce new features, but PCGH sources did not cite any examples.

Those new features could include an overhaul for how Arc users control graphics and schedule updates. Intel hasn't publicly revealed the new software, but Linus Tech Tips and Albert Thomas of Toms Hardware confirm that Intel told them it exists.

Intel Arc's current graphics control method, Arc Control, has faced heavy criticism, primarily because it's an overlay lacking an adjustable window rather than standalone software. Linus says the replacement app will consolidate Arc Controls' features into a single pane, while Thomas said it would satisfy users who don't like the overlay.

A recently leaked roadmap reveals Team Blue's plans to refresh the Alchemist lineup later this year before launching its successor – Battlemage – in 2024. While Alchemist only consists of entry-level and midrange GPUs, Battlemage might include Intel's first enthusiast-level card. The series is expected to improve ray tracing, memory, and machine learning.

Permalink to story.

 
Looking forward to a review. I'm still interested in seeing how good alchemist can be when the drivers are fully hammered out.
 
That's encouraging. I remember reading an official AMD dev response that they refused to work on DX9 drivers because they claimed nobody plays older titles. I should have bookmarked that, because it's so ridiculous to anyone that isn't some sort of cultist fanboy that it's almost unbelievable.

What's even worse is AMD somehow manages to break DXVK with their drivers, and DXVK 2.1 isn't even available for them yet (works fine on Nvidia). No new 6000 drivers in months either, and a good 7 or 8 months between recommended stable drivers that you could even use dxvk on at all prior to that.

All Intel has to do is beat that low, low AMD GPU bar and they can take second place in the discrete GPU sector. AMD clearly does not want market share.
 
Looking forward to a review. I'm still interested in seeing how good alchemist can be when the drivers are fully hammered out.
The problem is that it can take a decade or longer to get truly solid drivers for video cards. AMD and nVidia didn't get their drivers where they are in just a couple of years. Anyway, I hope Intel doesn't give up like they have so many times before.
 
I feel Intel's first stab at dedicated gaming graphics is a good one. While it falls short from high end gaming experience, I think the GPUs are still somewhat decent in terms of performance. Unfortunately, the fact that they finally released the GPUs very very late into the last gen GPU cycle means that it will end up competing with the current gen GPUs in terms of performance, and price point with previous gen GPUs. And this is where I feel the tyre falls off. So far, the availability of Intel GPU is looking very poor, and the price also looks unattractive. For example, if I look at pricing on Amazon, the A770 is in the price range can be higher than a RX 6700/6750 GPU.
 
So far, the availability of Intel GPU is looking very poor, and the price also looks unattractive. For example, if I look at pricing on Amazon, the A770 is in the price range can be higher than a RX 6700/6750 GPU.
Here in the UK, availability is pretty good, although only a relatively small number of retailers stock it. However, the price isn't very competitive - one can get a 16GB A770 for £390 but that would also get you a 12GB RX 6700 XT, which is a notably better card for gaming. Even RTX 3060s are cheaper than A770s.
 
Little by little. I'd consider this gen to be a dry run, but I'll be fascinated to see how this progresses over the next few months. If they can get the software up to a decent standard, Intel is only one good design win away from initiating market disruption.

The latest high end cards haven't sold well. I'd expect all eyes to be on the mid range market for the next couple of years, an opportunity Intel are reasonably well positioned to exploit should they continue to improve their execution. Hopefully will make up for the last few years of misery.
 
That's encouraging. I remember reading an official AMD dev response that they refused to work on DX9 drivers because they claimed nobody plays older titles. I should have bookmarked that, because it's so ridiculous to anyone that isn't some sort of cultist fanboy that it's almost unbelievable.

What's even worse is AMD somehow manages to break DXVK with their drivers, and DXVK 2.1 isn't even available for them yet (works fine on Nvidia). No new 6000 drivers in months either, and a good 7 or 8 months between recommended stable drivers that you could even use dxvk on at all prior to that.

All Intel has to do is beat that low, low AMD GPU bar and they can take second place in the discrete GPU sector. AMD clearly does not want market share.

What! So you want Intel to ignore the 80 to 88 percent of the market that Nvidia has and go after the rest of the market ..... hhmmmm does not seem like quite the money making idea that Intel has right now! I want them to start taking a bigger chunk of the higher end especially when their newer cards come out!
 
Back