The AMD Kabini Review: A4-5000 APU Tested

[strata8]

Still , they have an interesting power compsumption graph.

I refuse to believe that the ULV Pentium is consuming more power under load than the 35W Celeron. That's ridiculous.

 

[SPBHM]

I refuse to believe that the ULV Pentium is consuming more power under load than the 35W Celeron. That's ridiculous.

that's bizarre, it's the same thing, just lowe clock and voltage... it should use less power.

oh and even my b820 15" laptop (sandy bridge celeron at 1.7GHz, 35w) have lower power usage (13, 30, 45) than the ULV Pentium ivy bridge!?

 

[R0H1T]

I refuse to believe that the ULV Pentium is consuming more power under load than the 35W Celeron. That's ridiculous.

The difference is only 2~3W & can be counted within a margin of error, nothing unbelievable or overly dramatic scene being played out there !

edit : It could also be that the pentium is highly inefficient at high workload scenarios, seeing as how the TDP is only 17 Watt vs 35 Watt for the celeron.
http://www.cpu-world.com/Compare/43..._vs_Intel_Pentium_Dual-Core_Mobile_2117U.html

 

[ShintaiDK]

Still , they have an interesting power compsumption graph.

Not very useful tho due to the platform and vendor differences.

 

[SPBHM]

The difference is only 2~3W & can be counted within a margin of error, nothing unbelievable or overly dramatic scene being played out there !

well, it's supposed to be 17w vs 35w TDP, the difference should be much bigger if you were only comparing CPUs(+IGP)

 

[R0H1T]

well, it's supposed to be 17w vs 35w TDP, the difference should be much bigger if you were only comparing CPUs(+IGP)

It could be one of two things ~ techspot got a lemon for pentium OR they got wrong readings, in any case I'd take it as a major win for Kabini !

 

[Abwx]

I refuse to believe that the ULV Pentium is consuming more power under load than the 35W Celeron. That's ridiculous.

Believe what you wants if the numbers dont match your expectations,
theses are real consumers systems.

Acer Aspire E1 System Specs

• Intel Celeron 1020M (2.1GHz)
• Intel HM76
• 4GB DDR3 RAM
• Intel HD 2500 Graphics
• OCZ Vector 256GB SSD
• Windows 8 64-bit

Dell Inspiron 15 System Specs

• Intel Pentium 2117U (1.8GHz)
• Intel HM76
• 4GB DDR3 RAM
• Intel HD 2500 Graphics
• OCZ Vector 256GB SSD
• Windows 8 64-bit

http://www.techspot.com/review/671-amd-a4-5000-kabini/page2.html

 

[ShintaiDK]

Believe what you wants if the numbers dont match your expectations,
theses are real consumers systems.





http://www.techspot.com/review/671-amd-a4-5000-kabini/page2.html

Take 5 laptops from 5 different vendors. Same CPU/GPU. And see the differences in power consumption. Just like we see op to 30W difference just on desktop motherboards.

 

[strata8]

Take 5 laptops from 5 different vendors. Same CPU/GPU. And see the differences in power consumption. Just like we see op to 30W difference just on desktop motherboards.

I've looked at laptops with Intel's 17W parts and they all come in at 30-40W. Ditto for the 35W parts at 50-65W. I'm guessing it's just an error on Techspot's part.

 

[R0H1T]

I've looked at laptops with Intel's 17W parts and they all come in at 30-40W. Ditto for the 35W parts at 50-65W. I'm guessing it's just an error on Techspot's part.

OR it could be the actual power consumption of these low end & highly inefficient, compared to an i3/i5/i7 based part, chips from Intel ! Have you ever considered that possibility, not to mention the fact that TDP is not equal to power draw ?

 

[Enigmoid]

From LegitReviews:
http://www.legitreviews.com/article/2197/4/

Holy moly! I'm really curious to see the numbers for the 25W Kabini now.

I see a huge drop in platform consumption. I don't know what system they used to get 10+ watts idle WITHOUT the screen. i7 quads without the screen are going to use less power.

Note the delta between idle and full load is similar.

Nice power drop though. Doesn't help it though in larger notebooks or desktops (like brazos was used) where thermals are less of a concern.

 

[strata8]

I see a huge drop in platform consumption. I don't know what system they used to get 10+ watts idle WITHOUT the screen. i7 quads without the screen are going to use less power.

Note the delta between idle and full load is similar.

Nice power drop though. Doesn't help it though in larger notebooks or desktops (like brazos was used) where thermals are less of a concern.

Yes, it's odd. I've checked other reviews and an E-350 laptop usually consumes around 9-13W at idle with the screen on. So 13.6W with the screen disabled is surprising. Note the 9W idle figure at the top of the page for the reference laptop. Some of that can be attributed to that big 1080p display though.

AMD is attempting to satisfy that market with the 2Ghz quad core at 25W. It will sound good on paper at least

 

[Phynaz]

It could be one of two things ~ techspot got a lemon for pentium OR they got wrong readings, in any case I'd take it as a major win for Kabini !

Once one of their measurements are shown to wrong it calls into question all of their measurements.

 

[Abwx]

Sites numbers are more or less convergent for Kabini...

http://hothardware.com/Reviews/AMD-2013-ASeries-Kabini-and-Temash-Mobile-APUs/?page=7

 

[Vesku]

Interestingly enough Toms Hardware has been on a roll with decent reviews. First their article on Core 2 duo and quad versus current CPUs and now a Kabini review that actually includes a Pentium chip and not just the premium ULV Core i series.

http://www.tomshardware.com/reviews/kabini-a4-5000-review,3518.html

 

[leper84]

Thats funny because I just read the toms review and was thinking of how suckey it is.

A 35w (as in TWICE the wattage of Kabini) Pentium in a $500 laptop and a 17w (more like 20 something watt) i3 in an over $500 laptop with absolutely ZERO comparison the chip Kabini replaces or anything else in its price/power range.

I'd say the Anandtech review got it right. Toms has turned into trash.

 

[Vesku]

B960 (and its ilk) is one of the CPUs that none Temash Kabini will be competing with in the low end. The conclusion graph is a nice touch. I do have a minor quibble with saying the i3 3217U is in machines as low as $360. From what I can tell that's one Dell notebook on Newegg that has crap reviews and is marked down almost half off introductory pricing.

 

[AtenRa]

heh, anyone benched it against ATOM D2700 ??

 

[Abwx]

Cool , and so much for intel s allegedly 17w Cpu.
They perform well because they can go way beyond
the claimed TDP...

 

[Abwx]

heh, anyone benched it against ATOM D2700 ??

http://hothardware.com/Reviews/AMD-2013-ASeries-Kabini-and-Temash-Mobile-APUs/?page=5

 

[Phynaz]

Cool , and so much for intel s allegedly 17w Cpu.
They perform well because they can go way beyond
the claimed TDP...

They learned from AMD :biggrin:

 

[R0H1T]

Cool , and so much for intel s allegedly 17w Cpu.
They perform well because they can go way beyond
the claimed TDP...

This is been one of the major qualms I've had against Intel & their SDP'esque B$ because common knowledge says that AMD's superior GPU should make their offerings consistently stand upto Intel in the mobile dept but somehow there isn't much data to validate this point. Of course one of the reasons is the underwhelming number of models based on trinity & other similar APU's from AMD but alas one can't shoot their partners & hope to survive in this cut throat competitive PC industry, hence the dearth of quality products based on AMD chips is only now being felt/seen by the general public !

 

[Abwx]

The official TDP number is accurate for the pentium , interestingly ,
but not for the ULV i3 , that s how they artificialy create a higher
performing product marketing wise , just use the same 35W CPU ,
claim half the TDP but let them run at 30W for a few minutes ,
it really helps when they are rewiewed.....

 

[strata8]

B960 (and its ilk) is one of the CPUs that none Temash Kabini will be competing with in the low end. The conclusion graph is a nice touch. I do have a minor quibble with saying the i3 3217U is in machines as low as $360. From what I can tell that's one Dell notebook on Newegg that has crap reviews and is marked down almost half off introductory pricing.

AMD was pretty clear in their presentation about their positioning.

For reference:

Given that power graph above, I wouldn't be surprised if they're positioning the 25W Kabini against Intel's 17W i3, rather than the 35W i3 like you might first assume given the TDP Power consumption could be equal or even lower.

 

[R0H1T]

AMD was pretty clear in their presentation about their positioning.

Given that power graph above, I wouldn't be surprised if they're positioning the 25W Kabini against Intel's 17W i3, rather than the 35W i3 like you might first assume given the TDP Power consumption could be equal or even lower.

But that isn't how you do a notebook review, platform power & platform price is what matters more than each or any number of individual components put together with the former being more important of the two, however screen is the most important one as far as I'm concerned !