AMD Kills Monthly Graphics Driver Updates, Enthusiasts Out in The Cold?
Before we go down that road, we want to acknowledge that AMD's announcement just puts an official company stamp on a de facto situation. From 2010 through 2011, AMD released a new Catalyst driver every month like clock work. Starting last summer, however, AMD began having trouble with high-profile game releases that performed badly, were visually corrupted, or both. Rage was one high-profile example, but there have been launch-day issues with a number of other games, including Skyrim, Assassin's Creed, Arkham City, and Battlefield 3.
The company responded to these problems by quickly releasing out-of-band driver updates. These beta/preview/performance/pre-certified drivers (we've seen them called all of the above in the past nine months) have always had an unclear relationship with their validated brethren; when AMD released the 12.1 Catalyst series on January 25, it warned everyone playing Rage that they'd have a better experience if they stuck with the preview driver released three weeks previously.
ATI fanboys will protest that Nvidia also releases beta drivers on a regular basis, but the patch notes that accompany these releases show a very different pattern. Both companies devote a considerable amount of time to fixing multi-GPU bugs and edge configurations, but AMD's Catalyst 12.6 beta driver mentions random BSODs on the desktop, poor Crossfire scaling in Skyrim (forcing a return to 12.3 settings), and random hangs in Crysis 2 in DX9. AMD, in other words, is still working to resolve some important problems in games that launched more than six months ago. It's hard to put a positive spin on slower driver releases given just how often those releases are necessary.
AMD executives have openly announced that they intend to pursue a different competitive strategy -- one that doesn't rely on head-to-head competition with Intel. It's possible the company is mulling a different, APU-centric graphics strategy as well. Sunnyvale has never seen much net profit from its GPU activities, as shown below.
AMD has no plans to get out of the graphics business; GPU development is essential to the success of its APUs. With NV having recently one-upped AMD with Kepler, however, it's possible that Sunnyvale's CEO is taking a long, hard look at whether the discrete GPU business is a long-term winning bet. At the very least, we're certain Read wants to see the company's net profits improve on these parts. Companies like Imagination Technologies, meanwhile, have demonstrated that you can build incredibly successful mobile parts without offering desktop graphics.We took these questions to AMD, and the company affirmed that while it's not going to submit as many drivers for WHQL certification, it'll continue to release beta drivers as often as it needs to.
If AMD continues a rapid-fire beta driver release program, with only occasional WHQL updates, nothing much will change. If, on the other hand, this demonstrates how company resources are being directed elsewhere, we'd expect to see driver updates drop off, period. It all boils down to how much Rory Read wants to be in the high-end graphics / workstation business, versus how much he thinks the company would benefit from allocating more resources to APU software development. Short term, nothing is going to change, but keep an eye on AMD's Fusion summit and future executive statements.
|
Looks like I'll be jumping ship sooner rather than later, if AMD is set on getting rid of its GPU division. Too bad really, as I've always used ATI/ AMD videocards. |
|
BMAN, I don't think you can jump to that conclusion here and I'm not sure what this current announcement really means in the long run. We'll have to see how it pans out. |
|
Yea BMAN, the article has a tone as if he's narrating Rory Read's mind in third person but he isn't so don't take it so hard. A lot of people see what they want to see, beta drivers will release if a game is broken or severely hindered in some way, this is guaranteed or AMD has NO GPU DIVISION. People stop buying, GG. The idea behind this is that it's most likely in better interest for AMD to not be pressured in releasing a Catalyst Driver mandatorily each month. This allows programmers deeper access to manipulating drivers without having to make sure its functional before release. It's a larger window for improvement, allowing longer times before the need for time consuming extensive testing which would allow AMD to achieve greater stable performance. |
|
I think you mischaracterize my meaning, as well as the fundamental reason AMD brought in a new CEO. Rory Read is there to create shareholder value and maximize AMD's ROI. He has already discussed, in great detail, how the company is going to do to this around its future SoC's. No more pushing for new process nodes at the expense of optimized manufacturing costs, no more trying to challenge Intel in all markets. That's a given. That's what's known. Will AMD continue to have a graphics business? Obviously. But you don't have to have a business degree to see the problem their net margin in GPUs represents. Even during its *best* years, AMD's net profit in GPUs has been in the single digits. Why is that the case? I honestly don't know. The fact that 80-90% of the workstation market belongs to Nvidia is undoubtedly a major factor, but it's certainly not the only one. For AMD, the question isn't "Do we need a graphics strategy?" The question is "Do we need a workstation / enthusiast graphics strategy? I'm not suggesting that AMD is about to throw in the towel on graphics, but if you think this question hasn't been seriously considered, you're ignorant about the shake-ups that happen when a new executive brings in an entirely new executive team. Virtually no one that was running AMD two years ago is still with the company today. If AMD can't bring its net margins on discrete graphics, I'd expect to see the company moving to reposition itself in that market. |
stuff on Hot Hardware
Loading...
Loading...
Loading...
