*HP dv6z AMD Llano (6XXX series) Owners Lounge*

I still want to know if these can use normal, undoctored AMD drivers.... (Particularly if you're running with two GPUs...)

 

Should have no problem just downloading drivers from AMD and installing.

So no one has gotten a shipping notification yet? My estimated ship date is tomorrow but I am not getting my hopes up at this point.

 

AMD A8-Series A8-3530MX AM3530HLX43GX

 

I haven't heard anything since the delay email .

 

got a delay email today. pushed back until the end of july.

 

Still no tracking number, but I still have not received a delay email. Played on HP's website, and for the same config is now saying 08/11 estimated build date.

 

I'm in the exact same boat as you. I had orderd on 6/30 with an estimated build date of 7/13, but I still have recieved no update...

 

I ordered my Quad Edition on the 2th of July, and it shipped yesterday (13th), and expected to arrive 19th.

 

How does graphics switching work on this notebook? Is it automatic like Optimus or do you have to switch manually? Will a running accelerated program block it from switching? If you have to switch manually is there a button, or do you have to go into a menu?

Also, do the latest drivers allow you to select the discrete GPU only (instead of crossfire)?

 

It works like Optimus and there is an option to disable Crossfire but apparently it means it will only use the IGP.

My estimated ship date is supposed to be today but I haven't heard anything yet.

 

Intel or AMD? As far as I know, the current delays are for those ordering the AMD (the dv6zqe)

 

Anyone have links to any decent reviews or benchmarks with the A8-3530MX and 1GB discrete? At minimum 3DMark06 or 3DMark Vantage?

 

I for one would love to have some benchmarks to post. I don't believe any one has received the 3530mx yet however. There are a few around for the 3500 tho. There are links earlier in this thread.

 

There's a notebookcheck review of the 3410mx+6750m. You can probably estimate the performance of the 3530mx since the only difference is clock speed.

 

The problem with that review is the drivers or hardware or something isn't right. 7300 3DMark06? 1000 Vantage? What's up with that?

 

The 6490m and 6750m only have GDDR5

 

The drivers are really bad right now... Just look at the respective components. The 3530MX should be a little under Phenom II N930 level, the integrated 6620m (400 shaders @ 400mhz + crippled memory bandwidth) should be around 50-55% of the discrete 6750m (480@600mhz).

With 100% scaling, the 6755g2 should be 35% faster than the 6770m. This is very, very unlikely.

With 50% scaling, the 6755g2 should be around 10% faster than the 6770m.

With current scaling... which is like -20% and the added bonus of graphical artifacts, the 6755g2, well, sucks. The best case scenario with current drivers is like 15% scaling and considering how drivers will improve incrementally every month or two, you might be waiting quite a while before it matches even 6770m level (not accounting for possible CPU bottlenecks).

Also, didn't nvidia have a hybrid SLI gpu a couple years ago? (it was in the 9xxx series, iirc) How did that turn out?

 

Do we know though? Because you CAN'T do that with the Intel + AMD option

Well, that's good to know...and lame

Yikes...that's bad. And that's assuming you can actually install regular drivers on it without hacks

Yep. I guess they gave up on the idea, because I don't remember anything but that one product doing that.

 

kintetik said here he updated the drivers and he didn't make any mention of doing anything special. On the last-gen dv6z with switchable graphics you could just download and install.

 

Just got my delay notice. Pushed back to 07/28. Can't say I'm surprised. But I am disappointed.

 

If I configure today it says shipping like Aug 12. That's a long wait. If you order though do they charge you right away or only once your order has shipped?

 

I don't think it will be that long but I would assume July 20-something at the earliest.

 

I'm now thinking of cancelling the dv6zqe order and getting a sandy bridge quad core dv4t/dv6t/dv7t instead. There's now a 30% code for the sandy bridge versions.

A dv7t i7-2630qm with similar specs to my dv6zqe (including 1080p, blue ray, 6gb, 640gb 7200rpm) would run $1000~1100, depending on the options (the one I priced ran $1080 and my dv6zqe was 860ish, but I don't think the non-cpu/mobo/screen features were 100% identical). And of course, a 17 inch 1080p screen seems better than a 15 inch screen. But I think the dv6z offers better gaming (and maybe video/blue ray) on the battery, which is something I can use for travel.

I think a dv6t might run $950~1050 but did not price it.

 

I'm going to give it another week before I think about cancelling. I am curious to see what HP's sales figures for the first couple weeks has been. It doesn't make sense to me for them to offer something they can't provide at least a moderate amount of.

 

They charged me right away.

 

I am so tempted to cancel my NP8130, order this, at least I get my 21 day trial period anyhow right? I just don't want to be without a decent laptop for over a month, which is what shipping estimates are showing at the moment.

 

See, I'm on the fence with this one. If I ordered one, I'd pretty much max it, and even though it'll be quick, it won't match my M15x.. It might in GPU power in the future, or come close, but my i7... HT, that 8130 will be awesome, I bet, but the DV6z might actually game better on battery..

 

To be honest I'd go head first into this one if there weren't so many unknowns. From the DV6t owners you can't overclock and it downclocks on battery so not sure you can even get much performance on battery.

 

Yeah, seeing as we only have 2 people on the forum with these, and no one with MX series APU's, theres just alot of variables. I was also going to get one of these, but decided against until I heard more about the MX APU's and better drivers.

 

Same here, got pushed back to 7/28....maybe we can ask for some $$ credit?

 

It is worth a shot but if you ordered with the 30% coupon I wouldn't say it is likely you will get much, if anything. Assuming my order is delayed, it was supposed to ship yesterday and I haven't gotten a shipping or delay notice yet, I am gonna try.

 

yeah, I may still cancel my NP8130 though and see if I can just live with my M11x for a few months to let things settle a little bit until initial reviews and impressions on production units come through.

I would call. They may offer you some incentive.

 

I thinks worth a shot. I would be happy getting shipping upgraded to 2 day. I will call later today and post the results.

 

I just gave the DV6 a thorough look over at Best Buy. I have to say it's much better than I thought it would be. I'm now so close to going with this machine over the Sager now. Save me $400.

 

Does the dv6z has a numeric keypad?

Is its design is the exactly the same as dv6t? or different?

 

I wouldn't get an Intel one since you can't use AMD's drivers on them.

 

You can use AMD's drivers, you just need to get them from a 3rd party. There is an ATI drivers forum here on NotebookReview where you can download drivers.

 

With only having glanced at the DV6t, I believe the only differences are the guts. Someone else feel free to correct me if I'm wrong.

 

I think they are the same exact size but that the dv6z is .1lb heavier.

 

Not sure if anyone posted this yet or not but here's a good user's review of the Best Buy DV6-6135dx with A8-3500 CPU, 6750 GPU, 6GB DDR3, 1366x768 screen, blu-ray

HP Pavilion DV6-6135DX AMD Llano 3500M APU Review Dragons Gaming Central

Most interesting is that the 3DMark Vantage GPU score was P7349, for comparison stock 6770m is P5227, and 555m is P4821. 3DMark11 GPU was P2278, stock 6670m is P1340, 555m is P1856, but Crysis 2 benchmark was only 17.8fps average.

So either his results are wrong, the crossfire is working really well with the benchmark but not in Crysis 2, or Crysis 2 requires more hardware than I thought.

Of course the A8-3530mx is more powerful, but not by a whole lot more to make more than a couple fps difference.

 

those are pretty impressive results

but like I argued before actual games require a decent CPU to run well. hence GPU scores being top notch but actual game performance not being quite as great

 

Most games are GPU bound and CPU's these days, especially quad cores are more than adequate to power any GPU you put in there, especially in the mobile market. I had a desktop Core 2 Quad Q9550 running a GTX 460 eVGA superclocked GPU just fine. I'll dig up benchmarks but it shows even the lowest i3 mobile CPU doesn't crimp performance with a high performance GPU, in most games. GTA IV and Bad Company 2 are two games that come to mind that do better with a fast quad core, but that's not due to graphics, but mainly due to the additional processing for the game mechanics and such. Crysis 2 is definitely more GPU bound than CPU bound.

 

See the thing is you could make the very legitimate argument that the 3500M is equal to, or worse than a Core 2 Quad.

Of course games are primarily GPU dependent, no question about that. Upgrade that Core 2 Quad to a i7 and you will see a significant improvement in a game like Bad Company 2, and perhaps just as much in a single threaded game.

A comparison between different units of the same Intel Sandy Bridge architecture doesn't undermine my point.

To be honest, if the llano's CPU wasn't as bad as it is, I would get this machine because crossfire GPU power is better than a 6770m. But it seems to me that not only does it suffer from huge issues in games that don't support crossfire well and DirectX 9 games, but even in optimal conditions the CPU limits it to equal to the 6770m if its lucky.

 

Interesting set of benchmarks. I suspect asymmetric CrossFire is well optimized for 3DMark, particularly if you turn the resolution down and thus lower the memory bandwidth requirements.

Crysis 2 requires a 2GHz C2D. Unless it is perfectly threaded (which I very much doubt), the 3500M doesn't meet minimum requirements: it may be faster when all 4 cores are fully loaded, but that doesn't help the fact that its single and dual threaded performance is so slow.

One interesting fact: the OpenGL Cinebench score is consistent with only the IGP being used for OpenGL as mentioned on AMD's site. That is, the dv6t and dv6z suffer from exactly the same problem. Llano's IGP is more powerful than that of Sandy Bridge, but it's still far worse than the discreet GPU which you can't use for OpenGL on either model.

Sure... but the Q9550 is literally twice as powerful as Llano.

 

The comparison on Notebookcheck does undermine your point, jiggawhat, because as I said in the other thread the difference in even single-core performance between the i3-2310m and i7-2720qm and the A8-3530mx and i7-2630qm is virtually the same and yet there is absolutely zero difference between the former in their review in all except one game tested, and the difference there is only 4 FPS.

On higher settings the game is almost always GPU-bound. The i7-2630qm may get you 40FPS, or even more, on low settings but if you can tell the difference between 60FPS and 100FPS then go for it.

In regards to the benchmarks in the review, that is clearly a driver problem. Even the 5650m, which the IGP is about equal to, scores higher than that.

 

Can you back up that claim?

You're exaggerating. People typically play on settings where they get at least 30 fps. At these settings the game is mostly GPU bound yes, but a llano CPU vs a 2630qm can in some cases give you a 10+ fps boost that makes the game quite a lot smoother.

Given the lack of as comprehensive benchmarks for mobile parts, I can show you a desktop comparison from Anandtech's bench:

Athlon II X4 is what llano is based on. I chose a low-end sandy bridge that's dual core because the discrepancies between the two become so enormous at higher levels and you would probably dispute that.

15 more fps in sc2, 30 more in wow (yes wow is GPU limited mostly). The starcraft 2 benchmark in particular is one that does make a tangible, practical difference.

AnandTech - Bench - CPU

I took a top of the line athlon II and compared it to a 2 core bottom of the line sandy bridge. for mobile CPUs, where you're talking about an a8-3500m vs a 2630qm, I see no reason it should play out any more favorably for AMD.

 

Extrapolating from the A6-3410mx numbers on Notebookcheck, the A8-3530mx is 57% as fast as the i7-2630qm and the i3-2310m is 64% as fast as the i7-2720qm.

For one I hope you realize that, in single-core performance, the i3-2100 is faster than the i7-2630qm.

Secondly, those benchmarks were done with the Radeon HD 5870, which is at least 3 times more powerful than the 6770m. When the GPU is that powerful the CPU is going to make more of a difference. At the 67xxm level, the CPU is not going to make even close to that much of a difference.

Third, I said the SC2 was one of the games where the difference might be something noticeable. I don't play it, though, so I don't really care. Besides, it is in a small minority

 

See here: Benchmark Results: Crysis : Is Intel's Core i3-530 Fast Enough For Performance Gaming?

Look at the different benchmarks using desktop i3-530 in gaming. The A8-3530mx supposedly is more powerful than the i3-530 too. Results are more or less identical regardless of the CPU used and in worse case it's less than 10% affect on FPS performance.

 

Well, just got my delay notice. The updated date is the 19th, which I can live with.

 

Again, you are talking about a perfectly multithreaded scenario which is not the case with any game out there. In a dual-threaded scenario the i3-530 is more than twice as powerful as a 3500M (in fact, I suspect it will win out even in the perfectly multithreaded scenario). We have not seen a 3530MX, but I can guarantee you that it will be much, much worse when only two threads are needed.

You are right: most games have been GPU bound for a long, long time. But I think you are failing to appreciate just how far back in time you have to go to find a mainstream CPU as bad as Llano at tasks that full tax less than 3 threads. Any laptop Core2Duo from 4-5 years ago will utterly crush the 3500M in those circumstances. When your CPU is that bad, it will be the bottleneck in some games.