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

That is what they told me as well, except the one who said the 5650m and a couple who said they honestly didn't know. One of them sounded pretty knowledgeable, relative to his co-workers I guess, and said he was certain it was the 6770m. Maybe it is actually the 6770m and the Best Buy version is just different? That might explain why the Best Buy version is so much cheaper. It would also make more sense as the other discrete option is the same as the dv6t. I went ahead and ordered mine so I will let everyone know when I get it. It would be a nice surprise if it comes with the 6770m, because then the graphics performance could be up to 50% better than the dv6t, but not necessary for me.

• dark umber
• Genuine Windows 7 Home Premium 64-bit
• AMD Quad-Core A8-3530MX Accelerated Processor (2.6GHz/1.9GHz, 4MB L2 Cache)
• 1GB GDDR5 Radeon(TM) HD Dual Graphics [HDMI, VGA]
• FREE Upgrade to 6GB DDR3 System Memory (2 Dimm)
• 640GB 5400RPM Hard Drive with HP ProtectSmart Hard Drive Protection
• No Additional Office Software
• No additional security software
• 6-Cell Lithium-Ion Battery (standard) - Up to 5.5 hours of battery life +++
• 15.6" diagonal High Definition LED HP Brightview Display (1366x768)
• FREE Upgrade to Blu-ray player & SuperMulti DVD burner
• HP TrueVision HD Webcam with Integrated Digital Microphone and HP SimplePass Fingerprint Reader
• 802.11b/g/n WLAN and Bluetooth(R)
• Standard Keyboard
• HP Home & Home Office Store in-box envelope

$699.99 + $49 tax = $748.99 out the door with an estimated ship date of July 14th. Given my experience with HP, though, I will have it on or before then.

 

http://forum.notebookreview.com/har...grades/505702-amd-fusion-info-thread-105.html

 

ya HP customer service is 100% clueless, you gotta do your own research

 

Can you guys with the dv6 please try the new AMD Catalyst 11.7 drivers?
AMD quietly leaked them via some OpenCL debugger tool.

AMD gDEBugger | AMD Developer Central
First download, bottom of page.

 

Looks like the service guide is up on HP

http://h10032.www1.hp.com/ctg/Manual/c02842252.pdf

According to the guide it comes with the 6755G2. It gets pretty specific so I assume it is right.

@kinetik - Would you mind running Passmark's performance test and reporting the clock speeds? I am just curious because there are now 3 samples for the A4-3400M and they are about what you'd expect if it was running at the turbo speed.

 

Passmark is a random number generator. It ranks the A6-3400M (3075) as about equal to the Phenom II X4 B93, a 2.8GHz desktop chip with 6MB L3 cache (3083). Even if all 4 cores of the A6 were running at full Turbo, you'd still have cache-less K10.5 at 2.3GHz vs. full K10.5 at 2.8GHz. No way should they be that close. In case anyone was wondering, they can't get Intel CPUs right either. An i7-2720QM somehow beats the i5-2500K with 7220 vs. 7186 despite the latter being clocked a full 50% higher by default and 20% higher with Turbo Boost (assuming the 2720QM can max Turbo out which is a big assumption). It's just not a good benchmark.

That said, it would be funny if the effectiveness of Turbo Core varied drastically based on the workload -- people would have the option of buying two different types of Turbo (Boost and Core), neither of which is guaranteed to work at all times and which decide whether or not to work in different ways.

 

So did we determine whether turbo core was binary or not? Is it possible to hit clock frequencies between 1.5ghz and 2.4ghz for the A8-3400?

 

In this post, it is observed that it's binary. May be worth it to cross-check with various applications.

 

No but the first score that was posted for the A6-3400M was around 3600, I posted about it in the Fusion thread last week ( link), which is obviously wrong. Given the average of the 3 scores is now 3075, the average of the other 2 posted since then is about 2800. Given the scores of the Danube processors listed, that would be what you expect if all 4 cores were running at 2.3GHz.

IIRC, doesn't Intel say the Hyper-Threading gain can be up to 30%? And can desktop Sandy Bridge turbo all 4 cores to the given speed? Regardless, according the some benchmarks I have seen, the 2720qm isn't that far behind the 2500k. It wouldn't be out of the question for the 2720qm to be on par or even slightly ahead of the 2500k in some benchmarks.

 

Well, my credit card got declined apparently Dell.ca has not issued my full refund on the Vostro return yet. Good thing cause I plan on waiting for a DV7 Llano to get to Canada before school starts.

I'll just wait for you guys benches on the A8-3530MX + 1GB GGDR5 Radeon.

 

Order is in production. If my previous track record with HP holds, it will ship on the 7th or 8th and I will have it on the 11th or 12th. Here is to hoping.

 

Are you planning to upgrade the RAM yourself? The performance of the APU is strongly bound by memory bandwidth and that 6GB is not even proper dual-channel...

 

I figure if I got one, I'd grab an 8GB DDR3-1600 set for it, or if mobile Llano supports 1866, which I don't believe it does, that. Maybe if you can OC the RAM....

 

Yeah. I am, hopefully, gonna put 8GB of DDR3-1600. The maintenance guide for the dv6 says it only supports DDR3-1333 but the A8-3530MX is supposed to support DDR3-1600.

 

If the AMD could perform anything close to Intel then a lot more people would purchase it cause it is cheaper. However....

 

The A8-3530M is in between i3-2310M to i5-2410M in performance. Not exactly cheaper or faster just getting more for the money and personal preference. Take note the HP DVZ A8-3830MX HD 6770M 6GB 640GB Blu-ray WiFi (Ralink/Broadcom) comes to $700 and the DV6T i5-2410M HD 6770M (2GB) 6GB 640GB with Blu-ray WiFi (Intel) comes to the same price both with coupons.

With AMD you getting a quad-core CPU which will be slightly faster than i5-2410M in multi-core application but only as fast as a i3-2310M in dual and single core applications. You're getting a better IGP than the Intel HD 3000 and Dual Graphics HD which should really only make up for the performance lost in dual-core games against a i5-2410M but will be slightly faster in quad-core games. Other than the that it's not cheaper.

With Intel you getting a fast CPU for every use, single, dual and multi-core applications. You're getting 2GB HD 6770M which should be a bit better for scaling to a big screen TV or monitor (1080P or higher) and should be very close to the Dual Graphics AMD in dual-core games and only slight slower than the AMD in quad-core games dual to the AMD CPU being a real-quad. However, with the DV6T your getting a better Intel WiFi card and both a Intel and AMD logo stickers on your laptop.

It's really only a matter of preference now like AMD and NVIDIA videocards.

Llano will sell really well on lower end quad-core system without discrete videocards. A good example would be the A6-3400M with HD 6520G and A8-3500M with HD 6620G. Expect the A4-3300M with HD 6420G to take over the E-350 for the moment until till the E-360 comes out with better battery life.

 

1) 2GB makes no difference. Save your money
2) You're forgetting the 1080p screen. Don't skip that.

 

It's cheaper to get the i5 + 2gb 6770m regular edition than the i5 + 1gb 6770m select edition.

The 1gb 6770m isn't available on the regular edition.

 

I beg to differ, if you upscaling to a 1080P or higher in external the 2Gb will come into play also Eyefinity should be smoother with 2GB.

If you going 1080P screen you may as well get the 2GB 6770M. No need on Llano cause Dual Graphics reserves 512MB with system memory for the IGP thus making the total Dual Graphics card 1.5GB total.

 

Proof? Cinebench multithreaded benches show that the i5 is still faster in multithreaded apps even when assuming that the 3400m was running at 1.5ghz on all 4 cores the whole time and the 3530mx will be at 2.6ghz on all 4 cores.

 

Yeah but if the game has assets large enough to fill 2gb you'd be playing stuff at 10fps anyway.

There's no practical reason for 2gb vram, for gaming.

And since OpenGL doesn't work, most editing/modelling programs can't take advantage of it anyway.

 

I looked into the service manual of both the DV6 and DV7 Llano and it seems that HP is only shipping the highest models with a HD 6755G2 (HD 6770M), not is the HD 6750M can be overclocked with OEM drivers.

Another thing to note is that only 1x1 MiMO WiFi (Up to Intel 6230 2x2 with Intel CPU) with cards are supported with the Llano models and up to 8GB maximum memory (Intel is up to 16Gb with select CPU).

The memory does not bother me but the WiFi being only 1x1 MiMO maybe a problem for some, hopefully someone is able to mod the BIOS in the future to take the Intel WiFi cards off the white-list for AMD models.

...

 

Are you an intel representative? There is no reason to get an i5 over llano. For 99.9 percent of users, it doesn't matter if intel is faster than llano in obscure usage scenarios. How many times do you plan to use compression, 3d rendering, or encoding in the lifetime of the laptop? If those things are crucial to you, getting a laptop is probably not the wisest choice since a desktop will be more productive and cost effective. You see intel as a better value even tho you will rarely if ever use the full potential the CPU. Most users probably use their laptop to browse the web and there is no difference between intel or amd in this aspect. I know because I have both brands of laptops. You will however notice higher noise and temperature with the intel since llano is suppose to be extremely cool running from reports on this thread.

Utimately it all comes down to value. Savvy buyers would know what they need. Sadly the majority of people have fallen prey to intel's propaganda which touts certain advantages for intel even though most people will never use those capabilities.

 

I think your math is wrong somewhere. I got around 6300 in Cinebench multithreaded on the dv6-6135dx without a clean install. 6300/1.5 = 4200*2.6 = 10920. The average score for the i5-2410m is around 9500. If it can run all 4 cores at 2.6GHz, the A8-3530mx will be right around the i5-2520m in multithreaded performance.

But the dv6z can be purchased with 16GB. I think the manual must be wrong there

 

IMO you're an AMD apologist. look I want AMD to do well because without competition in x86 we the consumers will not benefit. but its clear to me that this generation AMD is well behind.

Llano is clearly inferior to the i5-2410m and above. Not to mention that you cite AMD's broken turbo, when Intel's turbo hardly works flawlessly either.

The benchmarks on anandtech clearly show the 3500M well behind sandy bridge, and even a slight boost in the 3530MX models won't do much at all to make up the very significant gap.

I'm convinced the dv6t is the better buy for maximum performance, and the dv6z is again a purely budget option (if you have an inflexible low budget). And I'm talking about gaming/graphics/etc. which is what most people buying the dv6 intend to do with it, otherwise you're blowing money on a less portable less battery life power hog.

 

The CPU is used in every single thing you do with your computer. How many times do you boot up your computer, install windows updates, do virus scans, install/uninstall a program? Even when you have an SSD, there are tasks that take multiple seconds, and those are CPU bound tasks.

I have used a low clocked C2D (1.73 Ghz UL30VT) before my current laptop with the i5-2540M that clocks to a beautiful 3.3 Ghz reliably every time thanks to turbo boost, and the performance difference is staggeringly obvious. I had the same exact SSD in both (moved my SSD from one to another) so that's not the reason for the speed, it's CPU only and it's noticable for me even when I'm not doing anything that would strain the CPU. It's just that much quicker to complete any tasks I do.

It's also plain BS to say Intels run hot, as my laptop runs cool and quiet. If you look at power consumption numbers, AMD consumes more power at load, and that's why Turbo Core was programmed to be very conservative at boosting the clock speeds of Llanos.

 

You're right, my bad. Was looking at single threaded numbers for some reason.

 

It's not necessarily inferior, just different. Intel and AMD made very distinct allocations of die space for Sandy Bridge and Llano. Intel chose to make a top-of-the-line x86 CPU combined with a GPU that has some great fixed function parts and unimpressive programmable performance. AMD chose to make an unimpressive x86 CPU with a very decent programmable GPU.

I have a somewhat unique perspective on this because for the past half year, I've been using a laptop with a dying 8600M GT that is no longer capable of hardware acceleration (it only works with the ye olde VGA drivers from the last millenium). When it first broke, I thought that I would have to buy a new laptop, but it turns out that there are remarkably few GPU-accelerated tasks out there. The only features that I really lost was the ability to play 3D games and use Windows Aero. Everything else works exactly as it did before, even things that you'd think would be reliant on the GPU (Youtube, Skype with video, Flash games, etc.). Of course, CPU utilization is higher, but by surprisingly little.

As a result of this, I'd take Intel's side here even for the casual user (I personally need a decent CPU for my work): the uses of a fast CPU may not be many, but the uses of a 400-core GPU are far fewer. Realistically, Llano's main advantage is cheap 3D gaming. A surprising number of people in the "What notebook should I buy?" section of this forum come there to ask what amounts to "Where can I find a gaming laptop for $500/$600/$700?" Before Llano, the only thing we could do was point them at outlet stores and such, but now there's a modern alternative.

Back on topic, I am not so sure whether it is worth it in the context of dv6t vs. dv6z. The gaming advantage is largely neutralized: the hybrid CrossFire may give the dv6z a 10-20% edge in some games, but it loses far more than that in CPU-bound ones. Is it worth the $200 or whatever to be stuck with CPU performance from 2-4 years ago, even for casual usage? I guess it depends on how much $200 means to you.

 

I have never denied any of this. I was just responding to the post that said the i5-2410m would be faster in multithreaded use than the A8-3530mx if all 4 cores could run at 2.6GHz. It is a fact that the A8-3530mx will be faster in anything that uses all 4 cores if it can run all 4 at 2.6GHz. I don't know how that makes me an AMD apologist.

The i5-2540m is about 300% faster than the SU7300, at the lowest. The i5-2410m is, at most assuming the A8-3530mx can't even keep one core that 2.6GHz, 50% faster than the A8-3530mx in single thread use. We are talking pretty huge differences here. Frankly, it is a terrible comparison.

You might want to actually back that up with something. According to every test I have seen, Llano under load consumes less than Intel. Even adding 10W for the A8-3530mx over the A8-3500m

Power Usage And Battery Life : The AMD A8-3500M APU Review: Llano Is Unleashed

AMD Fusion: A8-3500M A-Series Llano APU Review - HotHardware

Intel running hotter is just my personal experience, it doesn't reflect what may happen to everyone. But, from everything I have seen, Danube ran cooler than Arrandale in almost all cases and early reports, including in this thread, are that Llano runs cooler or just as cool as SB.

For one, the crossfire edge could be bigger than that. Secondly, there are far more games that will perform better with a better GPU than with a better CPU.

 

I think he was addressing the person who posted above you.

 

@ abaddon4180

You're right about the A8-3530MX scoring the same as the i5-2520M but from reviews on the desktop Llano a A8-3800 (2.9Ghz) beats the i3-2100K in mutl-core scores by a bit but looses in real life situations thus this makes me believe the mobile Llano would be the same. Synthetic scores wise the A8-3530MX should be up at the i5-2520M levels but in real life situations it can be where the i5-2410M is.

For me this good enough considering you get a better IGP but the fact HP locks their Llano models to just 1x1 MiMO really sucks. I'm pretty sure and hopeful something can be done about the BIOS WiFi white list.

Also, take into account that the A8-3530MX can be coupled with DDR3-1600 which should slightly put ahead of the i5-2410M in real life situations but you'll be sacrificing some battery life by doing so instead of using DDR3L-1333 low voltage 1.35v.

Also, putting in a 1x2 or 2x2 WiFi (if possible if BIOS mod) will also sacrifice some battery life on Llano HP DV7/DV6 models.

At the end of the day, I'd really like to get a Llano HP DV6/DV7 but for me the WiFi (lock/white list) is a big letdown right now on HP Llano models. I can get past the IDT soundcard as oppose to Realtek HD cause the DV6/DV7 has 4x speakers and Beats Audio but WiFi performance and range should be on top on everyone's priority life especially when buying a laptop. Another this to take into consideration for some users if the fact AMD has lackluster SATA drivers (slower than Intel) especially when it comes to trim for SSD's, this does not affect me in anyway because I will be using a 7200 RPM hard drive but may be a choice maker for some who must have SSD and need proper trim support with SATA drivers.

This makes things harder, all around better performance but giving up an essential all laptops need, a good WiFi card (more MiMO better range/throughput) that cannot be swapped by the end user.

Very tough right now, I'll just wait around for now.

 

I don't think so. His post said that whoever he was replying to cited Turbo Core, which the person above me didn't do.

 

Ul30Vt 1.73 GHz C2D gets 2107 from Cinebench R10 Single CPU benchmark while A8-3500M gets 2037. That's frankly pathetic for a non-netbook CPU released in 2011. Of course the 45W version will be a bit faster.

For the record, i5-2540M is around 5000 for single threaded Cinebench, so it's 250% faster. We can also interpolate from the 3500M performance 3530MX CPU to get (2037/1.5*1.9) = 2580 in Cinebench single thread, while the i5-2410M will get ardound 4500. That's more than 50% difference, more like 80%. I think that's a very significant difference as a lot of apps, even multithreaded ones, are usually waiting on one thread to finish (slowest path) even if the other threads are done.

From that same link:

That's just a weak A8-3500M vs. i5-2520M and they're about equal. The 3530MX is guaranteed to use more power than the slower clocked i5-2410M doing anything except hardcore 3D acceleration.

Anecdotal reports are not proof, I could say all AMD laptops I've used were POS (which isn't true because I haven't used any). However whether a single laptop ran hot or not depends on many factors in the construction of that individual laptop down to which thermal paste they used, so it's not fair to say which one runs cooler from a single personal example.

How many of those games would benefit from crossfire would narrow that down though.

 

20% is what you get if you add the FLOPs and assume that the 6750M and the most powerful Llano GPU combine with 90% efficiency. That is a really, really optimistic assumption -- it's about what ordinary CrossFire gets in the best case scenario and it's very far from what asymmetric CrossFire yields right now. AnandTech gets an improvement of 24% (which translates to a 71% combination efficiency) over the much weaker 6630M and that's ignoring the DX9 games (where it doesn't work at all) and pushing the settings up to the point where frame rates for some games drop below what is generally considered to be playable.

That depends on how bad the CPU is. It's true that the bottleneck in laptops has been the GPU for the past few years, but that's mostly because Intel makes much better laptop CPUs (relative to their desktop siblings) than what you get with either Nvidia's or AMD's GPUs (again, relative to the desktop variety). Depending on how Turbo Core works when the GPU is active, you could be looking at a best case scenario of a 1.9GHz K10.5 CPU. With something on the order of a 6770M, that will bottleneck you even in some of today's games and it will only get worse with time.

 

Llano's single thread performance is pathetic when compared with SB, or even Arrandale. That is not what we are debating here, though. We are debating on whether that is fast enough for the average user.

So the difference between the i5-2540m and SU7300 is still 3 times the difference between the i5-2410m and A8-3530mx, worst-case scenario. Like I said, it is a terrible comparison.

You never said that SB consumes less than Llano on average, you said that it consumes less under load. If you had I could hardly argue with that. I don't know how exactly Hot Hardware test the load consumption numbers but, even if you add 10W to them for the A8-3530mx over the A8-3500m, it is still going to be lower than the Intel systems.

Using the last-gen dv6 as an example, Danube ran cooler than Arrandale. Users with Arrandale-based dv6t's, even with i5 processors, were regularly reporting throttling and sometimes even shutdowns. Of the handful of users that owned a Danube-based dv6z, none reported any problems with heat. If you search these forums you will even find my posted screenshots of my temperatures in my dv6z review. It never went over 66C.

Besides, I am not telling me to buy Llano because it runs cooler. In fact, I am not telling people to buy Llano at all. I couldn't care less what other people buy. I am saying that I bought Llano because in my personal experience, AMD runs cooler than Intel.

20% over the 6770m is still a pretty big difference. I am not even assuming 90% efficiency, though, I am just assuming that the overclocking problems that AMD's dynamic switch brought one will be solved eventually. If the 6750m can overclock to within 5% of the overclocked 6770m like we have seen in other notebooks, then even something like 60-70% efficiency will push the 6755g2 ahead of the 6770m by 25%. 90% efficiency, which could eventually be reached, could mean the 6755g2 wins by almost 40%.

I am not sure this is true. Given the benchmarks on the 2630qm/GT555 vs. 2410m/GT555. Looking at the numbers for a heavily CPU dependent game like BC2, which is optimized to use up to 8 cores IIRC so the 2630qm should have a large advantage, the difference is almost nothing. The difference between the multithreaded performance of the i7-2630qm and i5-2410m is just as large as the single threaded difference between the A8-3530mx and i5-2410m, even assuming no turbo. And I think most games out there today are at least optimized for two cores so the difference will be even smaller. Even an N930, which is about where the A8-3530mx will be with no turbo, with a 5650m, which isn't even half as powerful as the 6755g2, gets 30FPS on high on BC2, compared to 42 for the i7-2630qm/GT555 combo above. In Starcraft 2, the story remains the same. The i7-2630qm/GT555 gets 7 more FPS than the P920/5650m on ultra settings. The max gain the i5-2410m/6770m will have over the A8-3530mx/6755g2, even in heavily CPU dependent games is going to be a handful of FPS, maybe less, on higher settings. On lower settings the difference may be larger but the human eye is not going to notice a difference between 90FPS and 120FPS.

 

Like I said, a non-ULV C2D (which is faster single threaded than a Llano) is probably fast enough for the average user, I'm not debating that. But people always want have faster and more future proof systems. Most people don't replace laptops every 2 years like you and I do. When it comes to replace their C2D laptops, people won't want something that's slower or about the same, they'll want something faster. Needs and wants are two separate things in this regard.

You're comparing two percentage differences. We're talking about a performance increase of 3000 Cinebench R10 benchmark points going from SU7300->i5-2540m vs. an increase of 2000 Cinebench R10 benchmark points going from A8-3530MX to i5-2410. It's certainly not 3 times difference between the two and certainly comparable.

Both are doing the same task, Intel uses more power and finishes the task faster and spends the rest of its time idle. AMD uses less power but for a longer time. When you add it up, Intel has consumed less energy than AMD. I shouldn't have said power, what I meant was energy. Hothardware power results are useless for comparing CPU's since they involve workloads that contain 3D gaming in the inclusion.

Or you can say HP dv6 is junk and not properly designed. I have 2 friends who do PC Tech support and repair and HP consumer laptops are the most commonly broken laptops that come to them. They both told me to avoid HP when buying a laptop. You don't hear about Asus or Sager laptops overheating and throttling.

http://www.pcworld.com/article/211402/reliability_and_service_laptops.html

 

I'd say HP DV6 refresh is better now since their taking the majority of features for last-gen Envy like CoolSense, Beats Audio, etc..

But if there were a choice I'd take a Lenovo or Toshiba Llano with HD 6750M over HP any day.

 

I don't think the average consumer will notice the speed difference between Llano and SB, let alone care about it. That is my whole point.

Sorry, don't know what I was even doing. I was at the beach all morning and I'm worn out. Still the i5-2540m is 2.5x faster than the SU7300, the i5-2410m is 1.8x faster than the A8-3530mx, worst-case. That is still a 150% increase versus an 80% increase. The former is obviously going to produce a more noticeable difference. Then we are back on the SSD discussion, though. For those who don't have an SSD, the difference is not going to be as noticeable because you are going to be waiting on the hard drive. Llano users are not likely to have an SSD.

Even if you want to say that Llano uses more energy, which is true, the difference is only 2.4W when web browsing and .1W in movie playback

I also think that HotHardware's load test involves stressing the processor. I don't see how including 3d gaming in the test would make it useless.

I can only compare apples to apples. There were no Asus or Sager laptops with Danube processors, as far as I know, but the dv6 was the one laptop that I have seen reviews for that had both available. Same laptop, presumably the same cooling system, same TDP, different temperatures.

And the HP support and reliability numbers have been discussed already in other topics. Looking at the chart you gave, HP and Dell are the worst. When the average consumer, usually someone who has no idea how to take care of a laptop, goes to Best Buy or Walmart or whatever to buy a laptop he or she will usually look for a brand they have heard of. HP and Dell are by far the two biggest manufacturers in the US. It makes sense that these people who buy for brand would have no idea what they are doing and, more often than not, cause the problems themselves. Acer is near the middle for the same reason. People who buy them are usually people with some knowledge of computers, enough to know it is not the brand that matters but the power to price ratio. Acer is usually ahead in said ratio.

 

Average consumer would not and should not upgrade a Core 2 Duo right now...

I say that because I have a Core 2 Duo laptop, and if I were the average consumer who only used the web and farmville, then I would probably use this same laptop for the next 10 years

 

You can't say that you're current system feels faster than your old one because the i5 is faster than a C2D. In everyday use there is no difference in terms of performance between the two. The difference is only noticeable in CPU intensive applications.

The fact that the SSD drive in both systems is the same doesn't mean a lot because the component in your current system (including the communication bus between the HDD and the CPU) are 3 years newer. Thus it is likely that you're SSD performs better in your new system than your old one. A fairer comparison is to test both the C2D and the i5 in your old (or current) laptop and see if there is a performance difference between the two. Unfortunately this can not be done because the C2D and Core i have different socket and hardware requirement

 

I take it that you've used the exact two systems and experienced no performance difference in everyday tasks? I used the UL30Vt for a year and there definitely is a difference between that and my current i5. CPU usage in the UL30Vt would frequently be pegged to 100% by the time I noticed the slowness and opened up task manager. I maybe saw it at 100% once in my current system.

The only difference would be the SATA interface. My current laptop has SATA6 while the Ul30VT had SATA3. However, since my SSD is SATA3, that has no effect. Both systems get 7.8 from WinSAT storage benchmark and also get similar numbers from Crystal disk mark. So you're incorrect.

The difference I see is due to much faster CPU and faster SSB (1333 vs 1066).

 

I've never liked overclocking laptop components, but even if you can overclock the 6770M and 6750M to the same degree (i.e. raise to core clock to roughly 850MHz which seems to be the average in this thread), it doesn't give you that much. I think you've misunderstood what I meant by efficiency: it's how much of the sum you actually get, not how much of the IGP is added. With the above overclock, you a 90% efficiency gives the 6755G2 an extra 30% performance, but a 70% efficiency gives you absolutely nothing (it's exactly 1.00 in that case).

If we define the efficiency your way (i.e. how much of the IGP gets added to the discreet card), then the average in the AnandTech results is only 39%. I have a very hard time seeing that go anywhere near 90% -- ACF is not the same thing as CF and its effectiveness decreases as the discrepancy of the GPUs increases. Of course, you can try it out and see; I could be wrong.

BC2 is more GPU-bound than CPU-bound. Also, comparing a 2410M to a 2630QM is not going to get you very far. The 2410M is well-suited to gaming: it is fast in applications with 1 or 2 threads which is what matters in most games. The 2630QM does have a bit of an edge because of the third core, but it's not large because you're still mostly limited by how fast you can run the main thread.

Llano is a rather different animal: it can never reach the 2410M in single or dual threaded performance. You can see the effect of that in the "Medium Detail" benchmarks.. Forget Sandy Bridge for now and compare the Llano + 6630M to the Acer 5551G (2.3GHz P520 + 5650M). Out of the 10 games tested, the inferior GPU wins in 3 (Civilization 5, Stalker, Starcraft II) and is much closer than you'd expect in another one (Mafia II). Having a quad-core is good for gaming, but not when your dual-threaded performance is that bad.

 

40% right off the bat is pretty good, and it will increase with better driver support. I also don't think that 90% efficiency, by your definition, is out of the question. That would only be about 70% by what I meant when I said it. Right now it is at 40% with the first drivers.

Like I said, I thought that BC2 was heavily multithreaded and thus the difference between the i7-2630qm and i5-2410m should be noticeable? I have never played BC2 so I don't know, that is just what I have heard.

If anything that proves my point about how little the CPU is going to matter, except in a handful of games. Also, the A8-3530mx is going be at least around 25% faster than the A8-3500m tested anyway.

But, lets look at the comparison between the P920/5650m and i5-2410m/6650m. The P920 only gets 93FPS on SC2 compared to 201 for the i5-2410m on low settings, a pretty huge difference (though some would argue that it wouldn't be noticeable). But, raise the settings to medium and you get 43 to 33, on high 28-26 and 17-17 on ultra. So, on low settings you may see a huge FPS difference, to some it would hardly be noticeable, but on higher settings the difference is going to be very little if it is there at all.

 

Well, with the first drivers it's 40% if you cherry-pick the games where it actually works. If you add all of the negative numbers it, I suspect it will be a whole lot less than that. In my experience, AMD is much better at making GPU hardware than they are at making GPU drivers. Maybe things will be different this time around, but I'm not holding my breath.

40% of the tested games is hardly "a handful".

That is correct. In fact, if you're lucky, it may even be able to Turbo properly with 2 cores even when the IGP is active. In that case, you're more or less set for today's games, though not for games in a year or two from now.

See, this is why I don't like notebookcheck for anything other than order-of-magnitude comparisons. The i5-2410m is faster than P920 under all circumstances (i.e. even when all 4 cores are needed) and the 6650M is at least 20% faster than the 5650M. So you have a faster CPU and a faster GPU... but the same exact score on Ultra. It makes no sense.

 

I think we can all agree with the last sentence to some extent but didn't the gains from symmetrical crossfire increase after being low right off of the bat? I don't think that ACF will work to the best of its ability for every game but I think it will, eventually and at worst, close the gap between the 6750m and 6770m.

I think it is only fair to point out that the i7-2630qm/GT540, which is actually slightly more powerful than the 6630m, loses to the P520/5650m in Civ 5 as well and the only game where the i7-2630qm takes a substantial lead is SC2, which I admit is one of the few games you could see large differences because of the relatively slow processor. I already said that the difference could be a handful to a dozen FPS and that is about what the i7-2630qm shows over the A8-3500m on medium settings. On higher details the difference will be smaller, though, and the difference between the A8-3530mx and i5-2410m, which we are discussing, isn't going to be as large as the difference between the i7-2630qm and A8-3500m.

I think 2 cores at 2.6GHz and two more at 1.9GHz is going to be fine for games two years from now. If not, then I doubt the i5-2410m will hold up too well either.

I don't like using Notebookcheck for their ranking system but their benchmarks are usually in-line with what I see elsewhere. Even in their other reviews featuring the i5-2410m and slightly better graphics, the 6630m or GT540m, the numbers hold true for SC2. Add at least 22% performance to the A8-3500m as well, possibly 45%, and the difference on high settings is going to be very little even in the most popular CPU-intensive game around. Sure, the i5-2410m could get you a hundred more FPS on low, but the human eye cannot tell the difference between 100 and 200 FPS.

 

Its available on HP's website right now...anyone thinking of purchasing or has already purchased?

I'm curious to see if the switchable graphics will work correctly on these units. And if there's a way to stay one card permanently...

 

Manual seems to indicate that HP is using ddr3-1333 memory, whereas I'm pretty sure that the AMD A chips do better graphics with faster memory (if my reading of the reviews out there is correct, there's a fps bump from 1333 to 1600)...

 

??????????

See how easy that was? Now how about you actually say something so I know whether or not you're actually providing an answer as opposed to just being annoying...

 

It has been available for a week now and he has already purchased the dv6-6135dx (A8-3500m/6750m) from Best Buy and I have already ordered the configuration in my sig. That is why he was questioning. I'm not sure why many sites just reported it yesterday