AMD’s Enduro Switchable Graphics Levels Up
by Jarred Walton on September 6, 2012 3:00 AM ESTOther Technical Details and Performance Expectations
So far we’ve discussed the past and near future of AMD’s Enduro/Switchable Graphics, but we haven’t gone into the technical aspects much. We’ve covered most of this previously (and neither AMD nor NVIDIA provide a ton of detail as to how precisely they’re doing the work), but there are a couple other tidbits we wanted to briefly discuss before wrapping up.
At a high level, all of the display outputs on a laptop now connect to the Intel iGPU, and AMD is able to route their content through the PCIe bus to the embedded graphics and out to the display. Nothing has really changed there; content is copied from the dGPU to the iGPU output in some fashion and you get the ability to switch seamlessly between the two GPUs. We also mentioned earlier that AMD has now removed the need for the active PCIe bus when the GPU is powered down, which drops power use of the dGPU from less than 100mW or so down to 0W.
One thing that hasn’t changed is AMD’s use of Link Adapter Mode (LDA) where NVIDIA uses Multi Adapter Mode, but we now have an explanation of why this difference exists. As far as we could tell, there’s not really an inherent superiority of either mode for general use. The primary reason AMD uses LDA is that they also have a chipset business, where NVIDIA has bowed out of making chipsets. Why this matters is that LDA is what facilitates AMD’s Dual Graphics (formerly Hybrid CrossFire)—the dGPU and the iGPU working together to render a scene. This is less important on Intel platforms, as AMD isn’t trying to do any cooperative rendering with Intel iGPUs; they potentially could in the future if desired, but that seems unlikely given the difficulty of getting even similar GPUs to work together. AMD also indicates that the use of LDA provides full support for Windows 8 Metro applications; I would assume NVIDIA also supports Metro apps, so unless that proves to not be the case (and we should know soon enough), other than Dual Graphics it appears that Enduro and Optimus are essentially at parity in terms of how they function, with software/drivers being the key differentiator.
Something else we’re still waiting to see is the packaging of the new Mobility Catalyst drivers. AMD didn’t provide us with the actual installation files—they installed them for us as they were still in a rather early state. That being the case, we aren’t sure if the Mobility Catalyst drivers for Enduro systems will feature totally independent drivers as far as Intel iGPUs are concerned, but that appears to be the case. If all goes as planned, you will be able to update your AMD dGPU drivers separately from your Intel iGPU drivers without any trouble once the Enduro Catalyst drivers start coming out.
Performance Expectations
AMD let us borrow a Sager notebook for a short time after the preview to test out the new “Enduro 5.5” drivers, and they also helped us install the drivers on a Clevo P170EM system from AVADirect. We’ll be providing a full review with performance data for the P170EM in the near future, but in the meantime we wanted to show off the Sager notebook as well as discuss performance expectations. Here's a rundown of the system specs for the Sager NP9150 along with some photos.
Sager NP9150 / Clevo P150EM Specifications | |
Processor |
Intel i7-3720QM (Quad-core 2.60-3.60GHz, 6MB L3, 22nm, 45W) |
Chipset | HM77 |
Memory | 8GB (2x4GB) DDR3-1600 |
Graphics |
Intel HD 4000 (16 EUs, up to 1250MHz) AMD Radeon HD 7970M 2GB GDDR5 (1280 cores @ 850MHz, 256-bit 4800MHz RAM) |
Display | 15.6" WLED Glossy 16:9 1080p (1920x1080) |
Storage | 180GB Intel 520 SSD |
Operating System | Windows 7 Home Premium 64-bit |
Price | $1919 as configured (9/05/2012) |
The Sager unit is their rebranded Clevo P150EM, and it has many of the same design issues that we’ve seen with Clevo in the past. The backlit keyboard with zoned lighting is a new addition, and they’ve tweaked the keyboard layout as well. Interestingly (and frustratingly), while they’ve mostly fixed my complaints with the 10-key layout, they went ahead and screwed up the main keyboard layout. The Windows key is now to the right of the spacebar, and there’s a second backslash key just right of the spacebar. I’m also not a fan of the tactile feedback from the keys, though it’s not terrible. Outside of the keyboard quirks and overly abundant use of plastic for the chassis, though, the performance is certainly there.
We ran through our current suite of games at the native 1080p on the P150EM with settings maxed out in most titles. Total War: Shogun 2 wouldn’t allow us to select Very High settings (a problem we’ve encountered on other systems in the past where the game incorrectly detects the amount of video memory and/or iGPU feature set rather than looking at the dGPU), but otherwise we get very respectable frame rates. Civilization V continues to be a bit sluggish at max settings (around 26FPS), but the brutal Battlefield 3 manages 36FPS and could very easily reach 40+ FPS if you disable 4xMSAA and just use FXAA. Those are the three lowest performing games we tested, with everything else running smoothly in the 45+ FPS range. A quick look at the last GTX 580M system we tested shows performance is better in over half of the games, and slightly lower in the other three titles. We’ll have a second look at the P170EM with a GTX 680M from AVADirect shortly after our full HD 7970M review, though, so stay tuned.
Finally, AMD did inform us that the current drivers aren’t fully optimized for performance (particularly with the 7970M), so we should hopefully see some gains with the final driver release—or if not then, the next release. Performance with GCN architecture desktop cards has been a little erratic since the launch, up until the latest Catalyst 12.7 drivers. I believe that the current beta drivers I’m using also predate 12.7 in some aspects (though they're version 9.0.0.0), so if that’s the case then the official release should clean things up quite a bit.
200 Comments
View All Comments
JarredWalton - Thursday, September 6, 2012 - link
Here's the thing: it takes time to correct course at a large company, and if management gets in the way it can take even longer. I don't get why people are outraged that Enduro 5.0 has issues when I said last year that Dynamic Switchable Graphics was broken and would remain so until AMD committed to regular driver updates.http://www.anandtech.com/show/4839/8
Did we ever come out with an article saying AMD was doing driver updates for Enduro? NO! So everyone that bought Enduro between last year and this was apparently willing to take a chance on AMD.
But now we have a demonstration of a fixed Enduro UI, and we have a firm commitment to two driver releases for Mobility Catalyst with Enduro support by the end of the year. And everyone is pissed about it and saying we didn't do our homework? Look at the introduction: "It has been just over a year since my last look at the technology, where things were so bad that I felt most users would be better off if they had only discrete AMD GPUs and no switchable graphics—or they could simply buy NVIDIA Optimus enabled laptops." I haven't changed my tune to recommending AMD Enduro over Optimus, but at least now there's a chance it will get there.
pablo906 - Thursday, September 6, 2012 - link
I'm with Jarred here. You guys are bringing the pitch forks to anandtech over a product they said not to buy. Does this make any sense to any of you outraged owners? If you have anger and frustration don't point it at anandtech for not talking about your issue with Optimus, take it out on the people who are continuing to let your issue go unresolved. AT hasn't looked at DSG in over a year, and when they did they reported it didn't work. They haven't been reviewing laptops with the technology at all since then and have not changed their stance from, this doesn't work - don't buy it. With that being the case when they get a product in that is having the performance and utilization issue, they start off saying this still doesn't work, but there is hope that it may soon. I don't know what more you can ask for other than perhaps an AT writer lighting themselves on fire outside of ATI in protest over the issue.TokamakH3 - Friday, September 7, 2012 - link
Can you point to the article where Jarred said faulty Enduro would cripple the 7970M's performance? I didn't care that Enduro wouldn't work right if I could shut it off. I didn't even consider the performance would be so horribly crippled by it, and AMD wouldn't even bother to even try to fix it. That's the worst part, 5 months of nothing.hulawafu77 - Thursday, September 6, 2012 - link
Well apparently I and many others were too stupid to not have read your previous articles to know AMD sold GCN 7970M units they were touting to be the fastest, most powerful graphics card even though last gen's Nvidia 580M were destroying it. We bought laptops since that is what AMD told us. buy this product, you get this performance.I don't see why it's out fault and why it was taking a chance.
I am grateful for Anandtech's diligence, but the anger I think here is more aimed at AMD than Anandtech. Some like myself just wish there was more push aimed at AMD to get this fixed now, rather than saying there is hope of rumored possible future update.
Seanzky - Thursday, September 6, 2012 - link
"I am grateful for Anandtech's diligence, but the anger I think here is more aimed at AMD than Anandtech. Some like myself just wish there was more push aimed at AMD to get this fixed now, rather than saying there is hope of rumored possible future update."This. At this point any false hopes of AMD fixing anything is better off whispered. We have reached out to every outlet we can think of. We have been ignored. All of a sudden AMD takes the platform provided by Anandtech and says that a fix will come soon? This is why I said this has AMD's hand all over it in my comment. It sounds to me like AMD is using Anandtech to further silence us. To silence us long enough until their next line of GPUs are out.
AMD to Anantech: "Just hang in there, tell them help is on the way."
Anandtech to us: "You hear that guys?! Hurray!"
You know what that sounds like to us abandoned customers? Like damage control, another false hope. Call me a conspiracy theorist, but I am merely relaying my real experience that others can definitely relate to. Put yourself in our shoes. No public statements. No acknowledgements. No signs. Nothing. With this article you want us to put down our "pitchforks" and torches? You want us to just go away and wait in our little corner? Damage. Control.
And if this was a known issue since day one, why is @catalystcreator say they're just "investigating" it. Why did Sager just investigate it also? Why did Sager go silent? Why did @catalystcreator go silent?
JarredWalton - Thursday, September 6, 2012 - link
Or how about this scenario, since we're just making stuff up.HD 7970 released, along with other GCN products. Management says, "Okay, desktop parts are out; get the mobile stuff shipping ASAP." Engineering gets the hardware ready and the driver guys say, "We're still working to address some concerns with drivers." They're told it's good enough for now and they start shipping. Then the forums light up with people saying, "OMG 7970M is terrible!" Management goes into a frenzy and tells the drivers people to fix the problem after marketing promises things will be fixed "soon". Unfortunately, marketing is writing checks that the drivers team can't cash.
Eventually, there's a meeting where the drivers team says they need to invest three or four solid months of work into fixing the underlying issues. Marketing knows that you can't come out with a statement like that, so they tell everyone to go silent and just wait for things to get better. Finally, the drivers are nearly ready and it just so happens that AnandTech has a 7970M review they're working on. It's a good opportunity to show that things are about to change, so they take it.
Heck, let me go one further. Here's an excerpt of what I sent to AMD (the whole email is about 4X as long):
-------------------
My conclusion for now, unless something can be done to address the situation in the next week or two, is that mobile gaming enthusiasts are far better off with NVIDIA GPUs, even if it means spending more (AVADirect charges $274 extra to upgrade from 7970M to GTX 680M). And actually, given the choice between a GTX 580M/675M Optimus and HD 7970M Switchable, it's $120 less at AVADirect to get the GTX 675M and you'd end up with a more consistent experience and regular driver updates.
To address the problems, AMD needs to get their switchable graphics solutions integrated into the regular driver updates, and users need to be able to update the AMD and Intel graphics drivers independently. Until that happens, I am extremely hesitant to recommend any AMD switchable graphics notebooks -- I'd rather have a discrete only HD 7970M or even 7730M/7750M/7770M with driver updates than to get switchable graphics without driver updates. This is pretty much the same thing I said a full year ago, and as far as I can tell virtually nothing has changed as far as the UI and driver updates.
-------------------
AMD's response was that they were working to fix the issue and they had pre-release drivers to show me, with the understanding that these drivers will be available sooner rather than later. And they showed me exactly that: updated drivers that they're saying will work for all Enduro (and PowerXpress 4.0 and later) laptops. They're not out now, but we have a deadline of October 31 with the potential to see them as early as this month. If AMD doesn't deliver, you can bet I'll have more to say on the subject.
vgray35@hotmail.com - Thursday, September 6, 2012 - link
Oh yes they were working on fixing the problem, and in the meantime behind the driver team's back, a covert sting operation was instigated to purge information in the public domain, that was focusing on the severity of the problem.Apparently while a fix is being engineered, it is not permissible for anybody to discuss the ramifications of the problem, and it is absolutely forbidden to discuss it in public.
Well I will discuss it where ever I please despite AMD position on this issue.
Seanzky - Friday, September 7, 2012 - link
Understood. You keeping on top of this is appreciated. Though, I no longer care about this per se, there are way too many over at NBR who deserve what they paid for. So, thank you.I hope you understand why my initial reaction was of AMD using you as a vehicle to further quiet us down. Many here have echoed the same things I've said and that was to either completely ignore us or to shut down threads where we bring up issues to resellers.
Vozier - Thursday, September 13, 2012 - link
Since we are imagining possible scenarios i might as well do..Only one thing is clear for now, and that is that ENDURO technology is responsible for underutilizing the dGPU and consequently rendering low FPS for some games and/or game maps and/or game configurations.
Now as a scientist i first analyze the empiric data and then try to formulate an hypothesis, and then test it to prove or discard it. Having avidly read many users info and comments, my current understanding of the problem is this:
I DO NOT think this is hardware issue. Why? well, the proof is everywhere, not only the 7970M performs flawlessly and very well in Alienware with enduro shut off, but also in clevo machines it performs perfect in some games and even in most games affected it performs normal in certain maps, multi or non multiplayer settings, etc, etc. The most common complaint is that FPS go up and down along with GPU utilization.
My understanding is that a hardware piece is inherently DUMB as a rock, and that its performance is only dummer as its operator (in this case Enduro) and the instructions it receives to operate.
Another evidence on this subject is that different catalyst driver versions do affect the performance, maybe not in the way we want, but it doesn just "stay the same", either worst, better or weird would be proper terms.
That said and taking the comments of the most knowledgeable people around, the issue IS driver related. The fact that Nvidia had very similar issues some years ago when they launched Optimus, just adds to this notion.
Another fact that can be interpreted in many ways is AMDs and most resellers silence about all this:
We can go all paranoid and think there is a huge conspiracy to shut us down and make this problem go away.
I can even picture some black clothed NVIDIA agent pouring a misterious liquid on the hardware or hacking AMDs drivers with a hidden flaw to cause all this.... the options are infinite...
Now, taking into account Anandtech's article, Sager's few deleted comments, mythlogic's comments, even short and very scarce AMD posts (Mark AMD), wich can be regarded as the most "official" position on the matter, the conclusion for me is this:
1) AMD has been slow to detect, understand, adress and moreso recognize this issue or its mere existence. This is not necessarily bad, only unfortunate (for them and us customers).
2) AMD took the usual attitude towards these kind of issues and remain silent, doing nothing at first hoping it was a mere configuration or driver installation issue and that other users would help the affected ones.
3) After seeing the issue was real, not fixable by users and configuration combos, not even reseller drivers, and that even some resellers were complaining to them and in open forums, AMD decided to put some people at work and asked their associates (i.e. sager and others) to back up their silence policy for the time being, releasing a few short phrases to gain time. "We are working on it"
you can argue this decision but i aint something so weird as it seems. I would say this happened only 1 month ago or so.
4) Viewing some light at the end of the tunnel AMD starts to release some info to their resellers and reviewers like Anandtech, in order to have them quiet and informed that they are "on to it". They still remain silent either because by now they might as well and because the fix is NOT ready or complete and they do not want to enter any debate without answers. This is happening now.
So as you see there is no need to be paranoid, there is much more simpler explanation to events.
Some would say AMD should have detected the issue before launching the 7000 cards, but seriously speaking this aint so true.
Why?
Well you need just read some of the first reviews that came out for 7970M, i dont think any of them were biased, they even mentioned enduro and its need to be improved. But if you test 20 or more games, you run only 1 or 2 minute benchmarks for them, you dont have someone playing hours and switching to multiplayer modes and what not. Also many games perform correctly so there really wasnt and still isnt an instant crash of sorts to detect a major issue. Also, the card works, even if it underperforms some times, so detecting this anomaly wasnt so easy to begin with. Not saying AMD is to be absolved of all charges, but i really dont see them as the villains many are picturing, they have been just too slow, too dumb and too arrogant, but then again who isnt sometimes?
Whats your fckng point you ask?????
Simple, the issue will be fixed to a good degree, not 100% probably, not for all games, but no card works flawlessly in all games, thats a whole different topic.
so have hope and dont crush the only option we have to avoid total monoploy in the graphics world.
What can we do then? sit back and wait?
NO, absolutely not, the ball is rolling and growing and we need to stop it soon or it will become an avalanche.
You can cry.
You can shout and spit.
You can test and inform your results.
You can post or just read.
Whatever you do will be more or less usefull and might help getting this boat to port.
best regards
Voz
transphasic - Sunday, September 9, 2012 - link
Yes. Agreed. I do not take issue with Anandtech at all, nor am I upset with them one bit here. I am also grateful at the work and info that Anandtech has been doing to inform us about this.In fact, in a way, this whole AMD Enduro fiasco has been a great help to Anandtech, because as of a month ago, I didn't even know that this website even existed until now. You can thank AMD in a backwards sort of way for helping me and probably a few others as well for making Anandtech's existence known to a great many more people.
Anandtech has helped me out with people like Jarred, who provided much-needed info, so thanks for that.
I AM upset and angry (obviously) at AMD for creating a product that we only now are regretting buying in the first place. If I/we had only known that this is what we would be going through, then NONE of us would be owning the 7970m, and we would all be siding with Nvidia, and talking about how great their 680m GPU is.