Trinity review

Page 34 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

AtenRa

Lifer
Feb 2, 2009
14,003
3,361
136
LOL! Your delusional. Neither IVB or Trinity are Good enough so at this moment in time . Idon't really care about whats available at this time . Now If Haswell is as good as some think . I may become interested in iGPU. But for now . Its good enough for everthing BUT gaming. I want the whole package . Lets see if haswell is that combination thats changes the compute world forever.

Heh, when your Intel is not in front you dont care about today's hardware. Im sorry to brake it to you but HW will have hard time overcome even Trinity's iGPU in games not to mention Kaveri's GCN.
 

Khato

Golden Member
Jul 15, 2001
1,248
321
136
Im sorry to brake it to you but HW will have hard time overcome even Trinity's iGPU in games not to mention Kaveri's GCN.

Eh, we'll have to wait and see. If HSW 'just' offers a similar improvement over IVB in 3dmark vantage as IVB did to SNB, then yeah, it'll remain inferior to Trinity's iGPU in some games and comparable to Kaveri in others. However there's always the possibility that Intel might have recognized their shortcomings in the SNB design and actually addressed them in HSW (feedback cycle would have been too short for IVB to properly fix major micro-architectural issues.) If that's the case then AMD could certainly be in trouble as Intel's problem with SNB and IVB is primarily that the iGPU is imbalanced/bottlenecked in some areas by comparison - on the workloads where they aren't hitting those issues IVB is already comparable to Trinity.

Have to remember that it's a lot easier for Intel to double or triple performance than it is for AMD. It's not likely that Intel will catch up from a pure GPU design standpoint, just as AMD isn't likely to catch up on the CPU side... but Intel's process advantage could easily make up for that deficiency once they get their architecture up to a certain point of diminishing returns.
 

AtenRa

Lifer
Feb 2, 2009
14,003
3,361
136
Eh, we'll have to wait and see. If HSW 'just' offers a similar improvement over IVB in 3dmark vantage as IVB did to SNB, then yeah, it'll remain inferior to Trinity's iGPU in some games and comparable to Kaveri in others. However there's always the possibility that Intel might have recognized their shortcomings in the SNB design and actually addressed them in HSW (feedback cycle would have been too short for IVB to properly fix major micro-architectural issues.) If that's the case then AMD could certainly be in trouble as Intel's problem with SNB and IVB is primarily that the iGPU is imbalanced/bottlenecked in some areas by comparison - on the workloads where they aren't hitting those issues IVB is already comparable to Trinity.

Have to remember that it's a lot easier for Intel to double or triple performance than it is for AMD. It's not likely that Intel will catch up from a pure GPU design standpoint, just as AMD isn't likely to catch up on the CPU side... but Intel's process advantage could easily make up for that deficiency once they get their architecture up to a certain point of diminishing returns.


GlobalFoundries installs gear for 20-nm TSVs


The systems should be in place and qualified by the end of July, with about half of them installed today, McCann said. The company aims to run its first 20 nm test wafers with TSVs in October and have data on packaged chips from its partners by the end of the year.

GlobalFoundries’ schedule calls for having reliability data in hand early next year. The data will be used to update the company’s process design kits so its customers can start their qualification tests in the first half of the year.

If all goes well, first commercial product runs of 20 and 28 nm wafers with TSVs can start in the second half of 2013 and ramp into full production in 2014, McCann said.

So far, three types of chip designs want to use TSVs. High-end mobile application processors will use TSVs to link to memories, high-end graphics and CPUs with use it to link to DRAMs and memory stacks that may or may not include any logic also will use TSVs, McCann said. All three classes could be in production in 2014, he said.
 

podspi

Golden Member
Jan 11, 2011
1,982
102
106
AtenRa, I'm not sure whether I should be impressed or depressed by that quote. On the one hand, that will put GF's 20nm coming in around the time of Intel's 14nm, which is more or less maintaining the status-quo. But my reading of your quote implies 28nm won't come online until 2014 either, which is terrible...
 

Khato

Golden Member
Jul 15, 2001
1,248
321
136

GlobalFoundries installs gear for 20-nm TSVs


The systems should be in place and qualified by the end of July, with about half of them installed today, McCann said. The company aims to run its first 20 nm test wafers with TSVs in October and have data on packaged chips from its partners by the end of the year.

GlobalFoundries’ schedule calls for having reliability data in hand early next year. The data will be used to update the company’s process design kits so its customers can start their qualification tests in the first half of the year.

If all goes well, first commercial product runs of 20 and 28 nm wafers with TSVs can start in the second half of 2013 and ramp into full production in 2014, McCann said.

The TSV portion is technically independent of the actual 20nm process. The more typical metric for progress on a new process being when the initial sram tapeout is - Intel showed 22nm sram at Fall IDF 2009 while Global Foundries announced the same at roughly the same time of year in 2011. So yes, best case execution for them would see full production in 2014... but since when have they been performing to best case execution? Not to mention, even if they do it'll be Global Foundries 20nm process going against Intel's 14nm.
 

AtenRa

Lifer
Feb 2, 2009
14,003
3,361
136
This thread is about Trinity and APUs, my quote above was related to APUs and what to expect from AMD and GloFo in the next one or two year in the APU front.
 

Riek

Senior member
Dec 16, 2008
409
14
76

Well the desktop model is also interesting.

HD4000 has about 160% more execution resources than the HD4000. Yet it delivers only about 70% more performance in games.

With such a core scaling, 24eu or 50% more than now will result in less than 20% performance increase... (heck even 40eu with the current scaling wouldn't make it much faster than llano on average..)
 
Last edited:

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,786
136
HD4000 has about 160% more execution resources than the HD4000. Yet it delivers only about 70% more performance in games.

My rule of thumb is that doubling memory bandwidth grants ~30% better performance, with doubling shaders bringing another 30%. 20% left over is everything else of course.

HD Graphics 2500 has 1 texture sampler with 6 EUs and HD Graphics 4000 has 2 texture samplers with 16 EUs. You can see my rule of thumb is pretty close, as increasing resources by 2.7x without memory bw increase should result in ~2x, except the texture sample isn't increased as much.

The performance gap in desktop between HD 4000 and HD 6550D is 30-40%. The HD 4000 in 3770K has 290GFlops while 6550D is at 480GFlops. That alone would account for only 15-20% and rest 10-15% is likely due to driver differences.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,786
136
For people arquing that HD4000 will perform the same from the High end CPUs to the lower end dual core i3 etc or in mobile,

There's actually a 17W using laptop that performs as well or even better than 35W parts.

http://www.laptopmag.com/review/laptops/asus_zenbook_prime_ux31a.aspx

World of Warcraft(Native Res)-
Macbook Air 11-inch(Core i5 2467M) : 28 fps
Asus UX31A(Ivy Bridge ULV): 22 fps
Gateway NV57H54u(Core i3 2350M): 15 fps
Acer Aspire 5755-6647(Core i5 2450M): 17 fps

So there is potential for much better performance, the others are likely bound by thermal limits and/or power supply issues. It's amazing how noone else can achieve it though.

 

386DX

Member
Feb 11, 2010
197
0
0
For people arquing that HD4000 will perform the same from the High end CPUs to the lower end dual core i3 etc or in mobile, have a look at the links bellow and compare Gaming benchmarks bettwen Intel Core i7 3720QM 45W HD4000 vs Intel i5 3427U 17W HD4000.

The AMD Trinity Review (A10-4600M): A New Hope

Intel Dual-Core Mobile Ivy Bridge Launch and i5-3427U Ultrabook Preview

Nobody said the HD4000 would perform the same from the high end to the low end. What has always been said is that the Intel IGP scales better from the high end to low because the number of EU stays the same while on AMD you lose both shaders and clock frequency. Put it this way...

Trinity GPU Desktop >>> Ivy GPU Desktop
Trinity GPU Mobile (High End) >> Ivy GPU Mobile (High End)
Trinity GPU Mobile (17W) ??? Ivy GPU Mobile (17W)

Where on the Desktop Trinity maybe 50% faster, on the high end mobile it maybe only be 30% faster. At the 17W Ultrabook level it could be 10% faster nobody knows right now... the Anand review even mentioned this:

Along with the lower TDPs of Trinity are lower CPU and GPU clocks, though, so while Trinity is clearly more potent for graphics applications at standard voltages, don’t expect the low voltage trinity parts to be quite as fast—and the A6-4455M loses a large number of Radeon cores, so its performance is really hard to guess at without hardware in hand.
What I find more interesting is the 2nd slowest of the four 17W IB process (i5-3427U) is faster then the fastest (A10-4600M) Trinity on the CPU side at both single and multi threaded apps. I really expected Trinity to at least be able to win the multi-threaded benchs. That doesn't bode well for the 17W Trinity with half the cores and slower clocks on the CPU side.
 

Vesku

Diamond Member
Aug 25, 2005
3,743
28
86
The top tier 35W Trinity A10 looks like it could almost function in the IB ultrabook chassis, thermal and power-wise. The 17W and 25W Trinity reviews will be quite an interesting read I think.
 

Joseph F

Diamond Member
Jul 12, 2010
3,522
2
0
When viewing a good quality Avi file , in that case Stevie Wonder Live at BBC,
CPU utilisation in the task manager with my old 1.6G P4 is lower than with my
DC T4400 2.2G and both have a discrete GPU.

Guess why....

I'm going to guess that there's a dedicated AMD GPU in the P4 laptop, and integrated graphics in the C2D laptop.
Honestly, that's irrelevant today, as both Intel and AMD now do a good job at hardware-accelerated video, AFAIK.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,786
136
Anand's power measurements are interesting. 25W cTDPup mode should bring additional 25% graphics performance and bring it mostly in range with 35W parts(though quads are still faster).

My guesses(iGPU):

-A8-4500M will end up 5-10% faster than A8-3500M.
-The LV A10-4655M will probably end up being ~5% slower than the A8-4500
-15-20% lower performance compared to A8-4500M for the 17W A6-4455M.

The LV version of Trinity should end up 10-20% faster in graphics than the 25W cTDPup version of Ivy Bridge. A6-4455M should be closer to the 17W version, but there's a chance it might still be 5-10% faster in average.
 

Abwx

Lifer
Apr 2, 2011
11,543
4,327
136

happysmiles

Senior member
May 1, 2012
340
0
0
they just announced Brazos 2.0 and....yep another paper launch.

Intel is killing them at computex
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |