Passmakr Baseline
http://semiaccurate.com/forums/showp...8&postcount=88
Thats all the benchmarks ATM.
Remember these are for a single module CPU,ie,it has one FPU and two integer units. The IGP is also probably the lowest one too.
Some Analysis
CB 11.5 MP scores are similar to an A4-3300M:
http://tech.sina.com.cn/n/2011-09-07...869667_2.shtml
http://www.notebookcheck.net/AMD-A-S...r.55741.0.html
The A6-3300M is a dual core with two FPU units and runs at between 1.9GHZ and 2.5GHZ as opposed to the single FPU of the 2.7GHZ to 3.2GHZ A4-4400M.
However,single thread scores are much higher.
The A4-4400M scores 37% more!!
OpenGL scores are slightly higher than for the A6-3400M.
This uses a 320 shader VLIW4 IGP. How many VLIW4 shaders the HD7540G uses is unknown.
The HD7540G scores P642 in 3DMark11 for the GPU.
This is higher than the HD6520G in the A6 CPUs:
http://www.notebookcheck.net/AMD-Rad...G.55734.0.html
The Passmark scores for the CPU and GPU are 25% and 50% faster than for the A4-3300M.
Last edited by CAT-THE-FIFTH; 27-04-2012 at 06:23 PM.
Biscuit (28-04-2012),DavidM (28-04-2012),watercooled (27-04-2012)
Next time I wont be arsed rehosting any benchmarks so quickly.
Last edited by CAT-THE-FIFTH; 27-04-2012 at 07:32 PM.
Just been assembling some furniture, took me a while to digest everything there. It does seem to be living up to expectations, I wonder how the desktop version will fare? Single-chip entry-level gaming PC is looking much more plausible now.
Cool
I seem to be more grumpy than usual - blooming ELK!!
Thanks CAT, huge amount of information to take in Looks like PD is a decent enhancement to BD, and the new graphics core (plus higher clockspeeds) will deliver a significant graphics boost over Llano too. Can't wait to see what 17W trinity can do: might be time to start thinking about a replacement for my Acer 3810TZ...
There's a lot to read on this thread and AMD will probly be on FM3+ by the time I've read it all, from what I can see of the performance boost on the A4 when moving from Bulldozer to Piledriver is upto 10%.
If that's correct should we expect to see a similar gain when looking at the FX8150 and FX8350 CPUs? Or is this likely to be a bigger performance gain?
There's been some conjecture that part of bulldozer's performance issue was down to a very high latency cache, and if AMD can find a way to reduce that latency without compromising final clock speeds then we could well see a higher performance gain going from BD to PD in discreet CPUs.
The key here is that FX CPUs have an L3 cache, whereas Trinity doesn't, so you can't make a direct comparison: but any performance gain Trinity shows over FX (the interesting comparison will be between top end desktop trinity and the FX4100/70) is likely to be amplified when PD CPUs hit the market. Also, it's hard to draw comparisons from a Trinity A4 as there is no single-module bulldozer available - as I said, the really interesting comparison will be between the dual-module (quad core) variants, so A10 v FX 4.
It seems AMD has gained some marketshare from Intel:
http://www.techpowerup.com/164915/AM...nse-in-Q1.html
http://uk.reuters.com/article/2012/0...83I1EM20120419
It seems they are doing better than expected - hopefully now they have been released from GF,they will be profitable for the rest of the year.
Last edited by CAT-THE-FIFTH; 30-04-2012 at 12:44 AM.
Details of the A85X:
http://www.fudzilla.com/home/item/26...-chipset-ready
It seems the Adobe Mercury Engine will be officially supported on Trinity:
http://www.theinquirer.net/inquirer/...s-trinity-gpus
Biscuit (30-04-2012)
Wow, Trinity is looking good. Guess I'll be getting it on release date (desktop variant can't be worse than the laptop variant now can it?)
There are currently 13 users browsing this thread. (0 members and 13 guests)