Phynaz
Lifer
- Mar 13, 2006
- 10,140
- 819
- 126
You should have effectively thought twice before posting
since i said nowhere that twice the thickness will amount
to twice the mass...
You certainly implied it.
When are you getting a PC?
You should have effectively thought twice before posting
since i said nowhere that twice the thickness will amount
to twice the mass...
Wrong , BT works close to 2.4GHz in MThreaded integer tasks
according to the very link you posted , hence optimistic or
wishfull misleading arguments..
Jaguar has 25% better integer IPC.
You're comparing a 2W chip intended for tablets to a laptop/convertible chip. You will never, ever see the the A4-5000 in the same small tablet form factors as BT-T will be in. Maybe you could compare the BT-T to the actual A4-1200.
BT should more adequately be compared to a 4C Temash , the SDPs are actualy close in most usages..
8W Temash consumes 8-10W under full load which you won't see from a 2-2.2W SDP Bay Trail-T. 8W Temash is for entry laptops, you might compare it to Bay Trail-M which should have a more comparable power envelope.
we havent really seen much abot z3770's platform power yet so I wouldn't be claiming anything so strongly but I do feel that the z3770 might use less power. Also we havent seen the a6-1450 in a tablet/power optimized device so also would discount temash.
You're comparing a 2W chip intended for tablets to a laptop/convertible chip. You will never, ever see the the A4-5000 in the same small tablet form factors as BT-T will be in. Maybe you could compare the BT-T to the actual A4-1200.
Comparing a 2W tablet SOC to a desktop or laptop chip is ridiculous. It's about how much CPU performance you can do in a tablet form factor, and the A4-5000 will not ever be in those form factors that BT-T will be in. I think the A4-5000 is a pretty good laptop/convertible chip, but to compare it to BT-T? Come on. The proper comparison to Bay Trail-T is the A4-1200 - and the BT-T has significantly faster CPU performance, easily 3 times as fast in that respect. The proper comparison for the A4-5000 is a core i7 quad core mobile chip which is a laptop and convertible chip - NOT the bay trail.
For the time being, the BT-T has the highest CPU performance applicable for that form factor and is faster than any ARM SOC. The GPU could be faster but it's an acceptable trade-off since nobody in their right freakin' mind plays crysis 3 on a 7 inch 2W tablet.
Intel has not released official TDP numbers yet. so unless anandtech gets to run actual power benchmarks with games and demanding apps we will never know. moreover Temash's A6-1450's max TDP is 8W. thats with Prime95 + Furmark running. for normal workloads its more like 3 - 5w. We are yet to see Baytrail Z3770 benchmarked by independent third party reviewers under different scenarios. With prime95 and furmark I am betting Intel will be closer to 6- 7 w. so lets reserve comments till we see independent reviews and power measurements.
Intel has not released official TDP numbers yet. so unless anandtech gets to run actual power benchmarks with games and demanding apps we will never know. moreover Temash's A6-1450's max TDP is 8W. thats with Prime95 + Furmark running. for normal workloads its more like 3 - 5w. We are yet to see Baytrail Z3770 benchmarked by independent third party reviewers under different scenarios. With prime95 and furmark I am betting Intel will be closer to 6- 7 w. so lets reserve comments till we see independent reviews and power measurements.
Doubtful that of the 11.6W delta between 10.6W maximum idle and 21.9W Prime95 + Furmark reported by notebookcheck's review of the A6-1450 - http://www.notebookcheck.net/Review-AMD-A6-1450-APU-Temash.92264.0.html - 3.6W or more was used by other components. aka, just because AMD claims it's an 8W TDP doesn't mean that such is true. Also note that the CPU was throttled down to 850 MHz in that fully loaded situation.
As for Baytrail-T power consumption while running a similar workload - feel free to make such wild speculations, but it's not going to change the fact that the SoC likely won't even touch 3W. HD4000 running furmark at 1.15 GHz only uses ~16W, divide that by four to account for Baytrail's smaller configuration, then in half to account for the lower frequency and you're already down to 2W (likely still an over-estimate as I'm not accounting for voltage scaling at all.) Now that's a very rough estimate, but it should be in the ballpark. Sure, 2W plus the reported 2.5W for multi-threaded cinebench would put it up to 4.5W, but that would require Intel to not reduce frequencies in order to keep total SoC power around the SDP.
Intel has not released official TDP numbers yet.
I wonder why they used a higher TDP chip with less performance? we are completely sure that BT-M and BT-D only have 4EU?
the acer v5-122p is a bad example for power, temash will eat more power but how much more is yet to be seen.
evidence. http://www.notebookcheck.net/Review-Acer-Aspire-E1-522-45004G50Mnkk-Notebook.101677.0.html
how can the a4-5000 with a 37Wh, a larger display and ~2x the tdp/higher clocks be on par with the a6-1450?
Because it's not? Their review of the A4-5000 itself using AMD reference platform - http://www.notebookcheck.net/Short-Review-AMD-A4-5000-APU-Kabini.93173.0.html - shows a 16.1W delta between idle and Prime95 + Furmark. That's quite a bit more than the 11W delta seen in that Acer review... and for some reason I doubt that AMD wouldn't have tweaked their reference platform sent out to reviewers to show their product in the best light possible.
Keep in mind that these are actual power measurements conducted by a third party. I find it amusing that the Baytrail power measurements are being called into question because they weren't conducted by a third party... and yet when third party Kabini/Temash measurements are provided that don't match with AMD's claimed TDP?
dont come at me with that crap, I linked to a real product and to demonstrate how another real product seems to have a higher powerdraw when it shouldnt.
acer v5-122p vs acer e1-522.
Why exactly shouldn't I point out logical inconsistencies? Especially since notebookcheck's review of the Acer A4-5000 shows the smallest delta between idle and load power consumption of the four A4-5000 systems that they have reviewed. (In addition to the reference system the other two - http://www.notebookcheck.net/Review-Packard-Bell-EasyNote-LE69KB-45004G50Mnsk-Notebook.99642.0.html - and - http://www.notebookcheck.net/Review-Toshiba-Satellite-C50D-A-10E-Notebook.100745.0.html - show delta power usages of 14W and 11.7W, respectively.)
There are a few options to explain the marked variations between the different samples. We can effectively rule out system variations seeing as how AMD's reference platform demonstrated the highest SoC power consumption under load. That leaves differences in software/firmware implementation resulting in the low samples being due to not actually running at full speed/load or huge variations in silicon quality in that some chips run at 10W TDP while others are around 15W... which is doubtful since I'd fully expect AMD to capitalize upon that rather than lumping them all into a single SKU.
(In addition to the reference system the other two - http://www.notebookcheck.net/Review-Packard-Bell-EasyNote-LE69KB-45004G50Mnsk-Notebook.99642.0.html - and - http://www.notebookcheck.net/Review-Toshiba-Satellite-C50D-A-10E-Notebook.100745.0.html - show delta power usages of 14W and 11.7W, respectively.)
so you are telling me nothing is wrong when the throttled a6-1450 uses as much power as the non-throttled a4-5000?Doubtful that of the 11.6W delta between 10.6W maximum idle and 21.9W Prime95 + Furmark reported by notebookcheck's review of the A6-1450 - http://www.notebookcheck.net/Review-...h.92264.0.html - 3.6W or more was used by other components. aka, just because AMD claims it's an 8W TDP doesn't mean that such is true. Also note that the CPU was throttled down to 850 MHz in that fully loaded situation.
I hope you are just thread crapping otherwise it is sad that you cannot read.So intel's reference platform using 2.5W under stress is a fabrication, and AMDs reference platform is a fabrication. I like where this is going.
SAMSUNG ELECTRONICS CO., LTD. 905S3G/906S3G/915S3G
Single Score 760 Multicore Score 1996
intel byt_t_ffrd10
Single Score 977 Multicore Score 2988
so you are telling me nothing is wrong when the throttled a6-1450 uses as much power as the non-throttled a4-5000?