The Notebook Review forums were hosted by TechTarget, who shut down them down on January 31, 2022. This static read-only archive was pulled by NBR forum users between January 20 and January 31, 2022, in an effort to make sure that the valuable technical information that had been posted on the forums is preserved. For current discussions, many NBR forum users moved over to NotebookTalk.net after the shutdown.
Problems? See this thread at archive.org.
← Previous pageNext page →

    The Ultimate AMD Trinity Notebook List

    Discussion in 'Hardware Components and Aftermarket Upgrades' started by davidricardo86, Jul 10, 2012.

  1. -ix-

    -ix- Newbie

    Reputations:
    41
    Messages:
    6
    Likes Received:
    0
    Trophy Points:
    5
    Thanks... learning a bit more I realised that now might be the worst rime investing in a new sleekbook, with both Haswell and (even more interesting) Kabini just around the corner.
     
  2. Gaugamela

    Gaugamela Notebook Consultant

    Reputations:
    10
    Messages:
    264
    Likes Received:
    4
    Trophy Points:
    31
    It seems that Richland truly is just a marginal improvement over Trinity. Still the Piledriver architecture, still the Northern Islands GPU architecture. :/
     
  3. Undyingghost

    Undyingghost Notebook Evangelist

    Reputations:
    78
    Messages:
    437
    Likes Received:
    33
    Trophy Points:
    41
    So wee need to wait "Steamroller" to extually get some improvements...thats what i thought.
     
  4. Atom Ant

    Atom Ant Hello, here I go again

    Reputations:
    1,340
    Messages:
    1,497
    Likes Received:
    272
    Trophy Points:
    101
    Yeah, but 1866MHz memory support for A10 Richland, that should bring some nice performance increase?
     
  5. Gaugamela

    Gaugamela Notebook Consultant

    Reputations:
    10
    Messages:
    264
    Likes Received:
    4
    Trophy Points:
    31
    But is it worth to wast 100$ to do such a marginal upgrade?
    This is pointless. IMO, just a way to save face for delays to Kaveri.
     
  6. Link4

    Link4 Notebook Deity

    Reputations:
    551
    Messages:
    709
    Likes Received:
    168
    Trophy Points:
    56
    Might be, since the iGPU is at least 20% faster and with DDR3 1866 might be even more. Its like buying a newer SSD, only slightly faster, but at least you can sell the old one on ebay.
     
  7. CharlieM76

    CharlieM76 Notebook Consultant

    Reputations:
    1
    Messages:
    102
    Likes Received:
    0
    Trophy Points:
    30
    Well, at least I have something to upgrade when it gets released! But I still want a 35W part in a 13 or 11.6 with anything higher than 1366x768.

    Yes, yes, I know.

    @Guagamela, you're absolutely right, it's not really that worthwhile of an upgrade considering I'm not unhappy with my current setup. This would just be a 'because i can' upgrade.
     
  8. davidricardo86

    davidricardo86 Notebook Deity

    Reputations:
    2,376
    Messages:
    1,774
    Likes Received:
    109
    Trophy Points:
    81
  9. Atom Ant

    Atom Ant Hello, here I go again

    Reputations:
    1,340
    Messages:
    1,497
    Likes Received:
    272
    Trophy Points:
    101
    According to AMD, the performance should be about 20 percent above the HD 7660G. Seeing your link David, all configurations used 4G DDR3-1600 (Dual Channel) Memory. So whith DDR3-1866MHz can we expect another 20% boost? If so, for 40% boost it is worthwhile to upgrade.
     
  10. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,877
    Trophy Points:
    931
    You might get a 5% boost due to 1866MHz RAM. Based of my testing of the Llano 6620G from 1333MHz to 1600MHz at same CAS resulted in about 6% on average performance boost (look at conclusion graph): http://forum.notebookreview.com/gam...g-benchmarked-various-ram-configurations.html

    And the clock speed is only a 5% boost from 685 to 720MHz, so I don't see massive improvements. Maybe 10% FPS improvement if we're lucky. Faster GPU clock % improvement does not translate directly 1:1 FPS improvement either. So 5% more clock could mean only 2-3% FPS improvement. So with faster RAM and clock speed improvement, you're looking at likely 5-8% FPS improvement. Not much.
     
  11. iamflang

    iamflang Notebook Guru

    Reputations:
    1
    Messages:
    67
    Likes Received:
    1
    Trophy Points:
    16
    Wow wingnut. You just shattered some of my hopes about Richland. I guess what remains to be seen is whether Richland really does allow some sort of overclocking if the cooling solution of the notebook can handle it, something I read might be the case in a few articles. If so, that might allow for more increase in performance. As it is, I'm happy with my A10 so no need to upgrade!
     
  12. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,877
    Trophy Points:
    931
    Well if the overclock potential is there then it might have some legs. They need to open up CPU overclocking like with the Llano. That being said, I can OC the 7660G in my A10-4600m and it makes games like Far Cry 3 playable 30FPS+ with low/med detail. BF3 is still unplayable, but most everything else is.
     
  13. iamflang

    iamflang Notebook Guru

    Reputations:
    1
    Messages:
    67
    Likes Received:
    1
    Trophy Points:
    16
    what do you use for that? amd overdrive?

    i was under the impression that overclocking was possible, but didn't offer much of a performance increase. I definitely have thermal headroom in this laptop-never has cleared 75C including extended gaming while charging the battery.

    also: good morning, sir!
     
  14. TommyB0y

    TommyB0y Notebook Deity

    Reputations:
    127
    Messages:
    1,501
    Likes Received:
    2
    Trophy Points:
    56
    Are you guys on dope? a 20% improvement of your GPU and 11% of your CPU is meaningless??? That is substantial. What are you expecting? Additionally, it was tweaked to get better turbo performance so you can run turbo longer or maybe you will be running 3GHz instead of 2.7 constant. That could make a huge difference that AMD would not test.

    The RAM should make a good difference, they likely used 1600 RAM because they took an off the shelf unit and put a Richland in it. Although, laptop performance 1866 RAM is not really common.
     
  15. davidricardo86

    davidricardo86 Notebook Deity

    Reputations:
    2,376
    Messages:
    1,774
    Likes Received:
    109
    Trophy Points:
    81
    AMD looks to HSA Foundation to avoid AMD64 mistakes

    Column Software developers hold the key to AMD's future
    By Lawrence Latif
    Fri Mar 08 2013, 15:01


     
  16. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,877
    Trophy Points:
    931
    Faster GPU clock with 1866RAM will equate to less than 10% performance improvement. CPU is only 8% improvement. If they can make it boost close to peak boost with all four cores, then that's another story, with this CPU I'll take any improvement I can get. But you need to be realistic. In reality for gaming, the FPS improvement will likely be 10% at best.
     
  17. TommyB0y

    TommyB0y Notebook Deity

    Reputations:
    127
    Messages:
    1,501
    Likes Received:
    2
    Trophy Points:
    56
    The GPU is still already better than the Trinity GPU, so I dont care what the difference between 1600 and 1866 will be. RIght off the bat, GPU is 20% better in 3dMark without the faster RAM apparantly. And Richalnd Piledriver CPU is 11% better in PassMark than the Trinity.
     
  18. iamflang

    iamflang Notebook Guru

    Reputations:
    1
    Messages:
    67
    Likes Received:
    1
    Trophy Points:
    16
    another interesting claim which would compel me to get a richland upgrade (in a year, when my warranty expires) is if they really have a 15% increase in battery life while streaming online video. that sort of battery life would push my 6475b into the 5 hour range, which would be quite nice.
     
  19. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,877
    Trophy Points:
    931
    Where do you get 20%? In the first footnote at AMD site: New AMD A-Series APU Offers Mobile PC Users Innovative Experiences, Elite Graphics Performance and Longer Battery Life

    it states:

    " The score for the 2012 AMD A10-4600M on 3DMark Vantage P benchmark was 4500 and the 2012 AMD A8-4555M was 2430 while the “Richland” 2013 AMD A10-5750M was 5000 and the 2013 AMD A8-5545M was 3930"

    Those are approximations and 5000 is about 10% more than 4500, not 20%. Although they do state they used 2x2GB DDR3-1600, so maybe another 5% with DDR3-1866, so approx 15%.

    One thing I've found is that 3DMark benchmark results equates to about half that % improvement in game FPS. So 15% 3DMark improvement is about 7-8% FPS improvement. That's like going from 30fps to 32fps. Not significant.

    However it depends on several things:

    (1) CPU boost, hopefully better than Trinity
    (2) GPU overclock headroom
    (3) Unlocked overclock of CPU or not.

    If we can have control over the P-state clocks and voltage like Llano it will have a significant impact on overall performance and improve cooling.
     
  20. Atom Ant

    Atom Ant Hello, here I go again

    Reputations:
    1,340
    Messages:
    1,497
    Likes Received:
    272
    Trophy Points:
    101
    Llano is different story, you cannot compare with Trinity GPU. Trinity GPU much stronger and more memory bandwidth hungry.

    TDP limit unlock would be welcome, otherwise with overclocking CPU you gonna kill GPU Turbo. Llano was not TDP limited, therefore both CPU and GPU cores could work together on their maximum clocks, even when CPU was overclocked.
     
  21. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,877
    Trophy Points:
    931
    It can be compared. I'm testing A10-4600m with different RAM like before. Well not as extensive, but 1333MHz vs 1600MHz at least. Not sure how it would be different, even Intel IGP results are similar.

    IIRC, Llano GPU couldn't be overclocked but CPU could. But yeah, it was really only limited thermally. Actually with my mods on my HP I was limited with voltage because I kept it cool, just couldn't bump voltage any higher.
     
  22. iamflang

    iamflang Notebook Guru

    Reputations:
    1
    Messages:
    67
    Likes Received:
    1
    Trophy Points:
    16
    I overclocked my 7660g to 800mhz using AMD Overdrive. I noticed it also locked in that frequency, didn't allow the frequency scaling that happens by default. i did not see a huge change in my 3d mark score. still testing games. how much fps change have you all seen overclocking your hardware? how high did you push the clocks?
     
  23. davidricardo86

    davidricardo86 Notebook Deity

    Reputations:
    2,376
    Messages:
    1,774
    Likes Received:
    109
    Trophy Points:
    81
  24. TommyB0y

    TommyB0y Notebook Deity

    Reputations:
    127
    Messages:
    1,501
    Likes Received:
    2
    Trophy Points:
    56
    Maybe I have just researched this a whole lot more than you, or spent 2 minutes googling it, because youre using a DX10 graphics bench instead of the 3DMark11.
    From AMD
    "Testing and projections develop by AMD Performance Labs. The score for the 2012 AMD A10-4600M on 3DMark 11 was 1150 and the 2012 AMD A8-4555M was 780 while the “Richland” 2013 AMD A10-5750M was 1400 and the AMD A8-5545M was 1100. PC configuration based on the “Pumori” reference design with the 2012 AMD A10-4600M with Radeon™ HD 7660G graphics, the 2012 AMD A8-4555M with AMD Radeon™ HD 7600G graphics, the 2013 AMD A10-5750M with AMD Radeon™ HD 8650G graphics and the 2013 AMD A8-5545M with AMD Radeon™ 8510G Graphics. All configurations use 4G DDR3-1600 (Dual Channel) Memory and Windows 7 Home Premium 64-bit. RIN-1"

    That is actually a 22% increase in 3DMark11, all configs using only 4GB of DDR3-1600.

    Now what say you, is it still "meaningless", and then throw in 1866 RAM and it should get even better.

    And here
    "Testing and projections develop by AMD Performance Labs. The score for the 2012 AMD A10-4600M on the PCMark 7 Overall benchmark was 1965 and the 2012 AMD A8-4555M was 1650, while the “Richland” 2013 AMD A10-5750M was 2175 and the 2013 AMD A8-5545M was 1850. PC configs based on the “Pumori” reference design with the 2012 AMD A10-4600M with Radeon™ HD 7660G graphics, the 2012 AMD A8-4555M with AMD Radeon™ HD 7600G graphics, the 2013 AMD A10-5750M with AMD Radeon™ HD 8650G graphics and the 2013 AMD A8-5545M with AMD Radeon™ 8510G Graphics. All configs used 4G DDR3-1600 (Dual Channel) Memory and Windows 7 Home Premium 64-bit. RIN-4"

    That is a 10.7% increase in PCMark overall score, not too bad, and also using DDR3-1600. Now if you can get 1866 RAM with CL9 that would be ideal. The Samsung 1600 1.35V 30nm 8GB kit seems to be able to run a nice overclock and low latency, so I might look into getting a set of those and reporgram to run 1.5V at 1866 CL9-9-9-27-36
     
  25. iamflang

    iamflang Notebook Guru

    Reputations:
    1
    Messages:
    67
    Likes Received:
    1
    Trophy Points:
    16
  26. davidricardo86

    davidricardo86 Notebook Deity

    Reputations:
    2,376
    Messages:
    1,774
    Likes Received:
    109
    Trophy Points:
    81
    This is clearly an old score, likely because of old AMD Catalyst video drivers. With 13.1 drivers now, Richland should also see a decent jump in 3DMark11 like Trinity did with more refined video drivers.

    I predict Richland will likely score 1500-1600 in 3DMark11, maybe even 1700 with 1866 RAM.
     
  27. Fat Dragon

    Fat Dragon Just this guy, you know?

    Reputations:
    1,736
    Messages:
    2,110
    Likes Received:
    305
    Trophy Points:
    101
    Any news on 35W Richland APUs in portable laptops? I'm expecting to buy in the summer (for US prices and customization options) so I can pass my Envy 14 on to my girlfriend, who is desperately in need of a halfway-decent laptop. Whether I get Richland, Haswell, or Kaveri (not out by summer, unfortunately), I want something in the 13" range, and since a 13" laptop could realistically use a 35W *PU, I'd just as soon get something with more power if they actually make it available.
     
  28. Atom Ant

    Atom Ant Hello, here I go again

    Reputations:
    1,340
    Messages:
    1,497
    Likes Received:
    272
    Trophy Points:
    101
    This how Trinity jumping with faster RAM;

    [​IMG]

    That is 12,4% difference between DDR3 1600 and 1866MHz RAM, in Games and applications! More here.
     
  29. -ix-

    -ix- Newbie

    Reputations:
    41
    Messages:
    6
    Likes Received:
    0
    Trophy Points:
    5
    I have read a number of reviews of the A8 4555M where a single RAM unit is said to severly limit the performance of the APU, but none where this is actually shown/tested.
    Are there such tests? If so, could someone point me in the right direction? Thanks!
     
  30. Fat Dragon

    Fat Dragon Just this guy, you know?

    Reputations:
    1,736
    Messages:
    2,110
    Likes Received:
    305
    Trophy Points:
    101
    There's evidence in this thread if you dig deep enough. I'm pretty confident HTWingNut has done some tests and reported on them, and there are probably links to other sources here as well. You're probably not going to find evidence collected specifically with the A8 4555M, but with other Trinity APUs, the results of which will mirror what you'd get with that particular processor.

    In these APUs, the GPU performance is limited by RAM speed, and going dual-channel and using higher-clocked RAM are the two biggest things you can do to improve performance. I don't remember the percentage, but the GPU performance drop going from dual-channel (2 DIMMs) to single-channel (1 DIMM) is big. If you are considering a Trinity or Richland APU for a machine that's going to be used for gaming, getting a single-channel-limited machine is a mistake.
     
  31. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,877
    Trophy Points:
    931
    That's good info. However keep in mind that's also desktop. Not saying there isn't a correlation, but desktop seems it would improve considerably over mobile, not to mention that's low CAS 1866. You can find 1600 @ CAS 11 predominantly, CAS 9 is available too. 1866 is likely CAS 13 or 11 if you get quality RAM. CAS doesn't have a huge effect, but it can affect by a few %. In any case about 10% is realistic it seems.
     
  32. cognus

    cognus Notebook Deity

    Reputations:
    186
    Messages:
    877
    Likes Received:
    18
    Trophy Points:
    31
  33. TommyB0y

    TommyB0y Notebook Deity

    Reputations:
    127
    Messages:
    1,501
    Likes Received:
    2
    Trophy Points:
    56
    I am hoping the mobile APUs scale better with RAM. Becuase it seems like the desktop Trinitys are hitting a wall around 15GB/s, compared to the mobile APU using the same rated RAM is only 11GB/s. So hopefully that barrier is close to 15GB/s on the mobile APU too, and that RAM speed increases will increment bandwidth capability nicely until we get to that wall.

    Are there bandwidth benches for 1066, 1333, 1600 RAM for mobile Trininty? to see how they scale, or are they just all at a 11GB/s barrier?

    Because RAM timings play a factor in the bandwidth, given the same RAM, at lower timings you could get more bandwdith if its not limited by something else in the CPU or memory controller. Timings scale my bandwidth really well on my Vishera FX CPU. So maybe Piledriver APUs will have better RAM scaling.
     
  34. Link4

    Link4 Notebook Deity

    Reputations:
    551
    Messages:
    709
    Likes Received:
    168
    Trophy Points:
    56
    Those Trinity memory scores are great, even though on desktop, but I was thinking about getting the 8GB Corsair DDR3-1866 kit anyways, so hope I can get my hands n Richland soon. (someone sell it on ebay already!)
     
  35. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,877
    Trophy Points:
    931
    Here's my results of AMD Trinity A10-4600m with various RAM configurations: http://forum.notebookreview.com/har...0g-igp-tested-various-ram-configurations.html

    Summary chart is here:

    [​IMG]


    Summary:

    RAM at same CAS between 1333MHz CAS9 and 1600MHz CAS9 is about 7% improvement
    RAM with tighter CAS timings 9 vs 11 resulted in about 4% improvement
    RAM running Single Channel vs. Dual Channel results in about 30% GPU performance degradation
    RAM running 1066MHz vs 1600MHz results in about 9% improvement for each 266MHz increment.
     
  36. Link4

    Link4 Notebook Deity

    Reputations:
    551
    Messages:
    709
    Likes Received:
    168
    Trophy Points:
    56
    Now which benchmarks did you use, and if not 3DMark11 can you run that?

    Edit: Never mind, and if you have time can you test World of Tanks, it's very demanding at 1080p or high settings, and its a free download so you don't have to buy anything.
     
  37. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,877
    Trophy Points:
    931
    Click the link.
     
  38. Atom Ant

    Atom Ant Hello, here I go again

    Reputations:
    1,340
    Messages:
    1,497
    Likes Received:
    272
    Trophy Points:
    101
    Good to see even CAS latency have so good affect for performance, so now I want 1866MHz RAM with CL9 timing. Anyone has idea which the current most customizable Memory for TaiphoonBurner? I actually have Adata 1600 CL9 1.35V Ram, maybe could take 1866MHz CL9 near 1.5V?
     
  39. TommyB0y

    TommyB0y Notebook Deity

    Reputations:
    127
    Messages:
    1,501
    Likes Received:
    2
    Trophy Points:
    56
    That one or the Samsung 30nm 1.35V 4GBx2 1600 Kit. I am really hoping AMD improved the memory controller, so that it gets a good jump in bandwidth.
     
  40. Atom Ant

    Atom Ant Hello, here I go again

    Reputations:
    1,340
    Messages:
    1,497
    Likes Received:
    272
    Trophy Points:
    101
    I just checked my ADATA, but write protected :(. Any link to that Samsung Memory? It seems Memory prices going high, so I might have to act quick...
     
  41. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,877
    Trophy Points:
    931
    Samsung 1.35v was able to go to 2133 in my W110ER with slightly looser timings (CAS 12). It's one hell of an overclocking RAM chip. I was also surprised at the effect of CAS, a bit more than I expected.

    @Link4 - The IGP really isn't a 1080p chip. It's really deigned to run games at 720p due to the limited bandwidth and processing power. I did some testing at 720p and 900p for a bunch of other games with stock RAM, but haven't published them yet, been so busy with so many other things. Will do as soon as I can. If I have time I'll give World of Tanks a try. Probably at 720p and 900p and if it looks like it stands a chance, hook up to my 1920x1200 monitor, but I doubt it.
     
  42. Loney111111

    Loney111111 Notebook Deity

    Reputations:
    396
    Messages:
    828
    Likes Received:
    28
    Trophy Points:
    41
    Richland has a 1866 MHz memory controller. There are rumors that Kaveri may have GDDR5 support, and there is a chance that the mobile version may support 2133 MHz RAM (hopefully).
     
  43. TommyB0y

    TommyB0y Notebook Deity

    Reputations:
    127
    Messages:
    1,501
    Likes Received:
    2
    Trophy Points:
    56
    Yes, it has the 1866 controller, but the 1600 controller in dual channel is not even hitting the theoretical bandwidth of a single channel of 1600 RAM. So hopefully the 1866 controller is not just going to run up against the same bottleneck that is showing such little performance gain from faster RAM past 1333 in dual channel.

    The Samsung RAM is not such a bad price on Amazon
    Amazon.com: Samsung Electronics - 30nm SODIMM 8 Dual Channel Kit DDR3 1600 (PC3 12800) 204-Pin DDR3 SO-DIMM MV-3T4G3D/US: Electronics

    And it looks like guys on the boards have already been discussing it for a while, including Meaker who has a GX60 and active on that thread.
    http://forum.notebookreview.com/har...extreme-boost-1600-2133-beyond-jedec-xmp.html
     
  44. Loney111111

    Loney111111 Notebook Deity

    Reputations:
    396
    Messages:
    828
    Likes Received:
    28
    Trophy Points:
    41
    Does AMD's APUs recognize XMP profiles?
     
  45. Atom Ant

    Atom Ant Hello, here I go again

    Reputations:
    1,340
    Messages:
    1,497
    Likes Received:
    272
    Trophy Points:
    101
    Good point, so what would be the normal bandwidth with Dual-1600MHz? My A10 has made 10384MB/sec, while my Llano only 8173MB/sec with 1600MHz CL9 RAM.
     
  46. Link4

    Link4 Notebook Deity

    Reputations:
    551
    Messages:
    709
    Likes Received:
    168
    Trophy Points:
    56
    I run World of tanks on 1080p and almost all settings maxed out but and almost always stay above 20fps (its not too laggy in that game, ping is the real issue sometimes, and with minimums at 20 its really smooth) but thats with A10-4600m + HD7730m, so not iGPU, the only way i can test the iGPU is when i run off of a battery but I am not sure if I get max performance out of the APU then. When I run in iGPU on battery it seams like the CPU starts struggling and the framerates drop to 15 (or might be because of the weaker iGPU, but it seams like the 7730 is kept down by the CPU)
    which leads me to think that te biggest issue is the CPU and iGPU in the APU sharing resources causing it. Now I'm not sure if thats a TDP or a memory bandwidth issue (or a memory buffer issue as my BIOS only alows my iGPU to use 512MB of system RAM, which i tried to increase but couldn't find an option in the BIOS). Anyways if it's the latter I am really going to have to upgrade my memory and get something better than the DDR3-1600 CAS11, hopefully DDR3-1866 Cas10 if i get my hands on Richland.

    Also Assassin's Creed 3 runs very smooth on my system when plugged in as when I first started the game it put everything on medium settings and claimed that those were the recommended settings. I then increased the resolution to 1080P (native res) and then increased everything to MAX and surprisingly I din't notice any lag at all, even when running on rooftops and getting shot by British soldiers. And what I even noticed recently is that my Catalyst Control Center has a High Quality
    option turned on which forces 4XAA on everything and its still running without lag. I tried to tun in on battery on the same settings which means the dGPU was turned off and suprisingly there was very little lag. Now it interests me what can I get with good RAM in my system when neither he CPU nor the iGPU are starved.

    Another very demanding free game I have is Vindictus and when I first downloaded and ran it on my new laptop I was surprised that going from no AA to 4XAA and then back to 2XAA didn't make a difference and thats when I checked the CCC and found the forced on 4XAA option. This game unlike Wold of Tanks is more GPU demanding compared to World of Tanks which i believe is more CPU limited. There is a big difference in going from soloing dungeons to running 8-24 person raids and areas with a ton of mobs and the framerate varies in this game a lot and is't smooth like WoT where you can have minimum 15 fps and everything is fine and game-play is unaffected. As explosions happen your fps drops from the average of 30 to under 10 and then its like watching a slideshow and it gets so bad you don't even see your attacks happening but only monsters flying everywhere and your character sliding on the floor.
     
  47. TommyB0y

    TommyB0y Notebook Deity

    Reputations:
    127
    Messages:
    1,501
    Likes Received:
    2
    Trophy Points:
    56
    theoretical bandwidth of 1600 single channel is 12.8GB/s, why its known as PC3-12800, which is MB/s. Dual channel then would be double that, and desktop FX CPUs can get like 17-21GB/s, but the desktop APUs seem limited to around 15GB/s where the mobile APUs around 11GB/s. All on 1600 RAM. So, if the new controller actually bumps the 1600 RAM bandwidth up closer to 15GB/s that could be a big part of the iGPU gains of 20%. But hopefully, it actually scales so that the test done with 1600 RAM are not as good as with 1866 RAM, or you get more than 5% improvement anyway.
     
  48. darnok44

    darnok44 Notebook Consultant

    Reputations:
    177
    Messages:
    285
    Likes Received:
    87
    Trophy Points:
    41
  49. HTWingNut

    HTWingNut Potato

    Reputations:
    21,580
    Messages:
    35,370
    Likes Received:
    9,877
    Trophy Points:
    931
    @Link4 - you can choose to use the IGP whether plugged in or not through the control panel. You don't have to unplug to do that.
     
  50. Atom Ant

    Atom Ant Hello, here I go again

    Reputations:
    1,340
    Messages:
    1,497
    Likes Received:
    272
    Trophy Points:
    101
← Previous pageNext page →