Intel Atom Z2580 AnTuTu score drops 20% after revision to site

Intel Atom Z2580 AnTuTu score drops 20% after revision to site
Earlier in July, a benchmark test of the Intel Atom Z2580 chip resulted in a very impressive score. The test on the AnTuTu site revealed that the Intel silicon had beat out processors from Qualcomm, NVIDIA, and Samsung. But as it turned out, the score was somewhat misleading. Consulting firm BDTI pointed out that the Intel chip was not executing all of the instructions. Because of the ICC compiler used for Intel processors, the RAM test was incomplete which artificially raised the score. Other processors use the GCC compiler.

The AnTuTu site has since been revised and the benchmark score for the Intel Atom Z2580 is now 20% off with the RAM score itself cut in half. Scores for other processors remained the same. The 20% cut in the Z2580's score now puts the Samsung Exynos 5 Octa on top of the Intel silicon based on benchmarking the Exynos 5 Octa powered Samsung Galaxy S4 against the Intel Atom Z2580 flavored Lenovo K900.

Things could get wild next month when a major revision is scheduled for the AnTuTu testing standards.


source: EETimes via SlashGear

FEATURED VIDEO

27 Comments

1. drum009

Posts: 33; Member since: Jan 23, 2012

whatever the Hell!!!!

19. ihatesmartphone unregistered

Don't care much about benchmark. lol XD

20. TheLolGuy

Posts: 483; Member since: Mar 05, 2013

After stepping on AMD's toes with anti-business tactics to stay ahead with an inferior product, I no longer believe any numbers or promises they make/release. They have a giant operation and about ~15 times the R&D expense of AMD (or so I've read) so I don't doubt they'll bring something competitive. Just don't read too much into their lies.

21. ihatesmartphone unregistered

So true, they have long monopoly in pc markets (same like Microsoft, not so many people are using linux or mac os, lol) and right now they want to enter in mobile business, so good luck for them because... you are entering in very competitive market!! :D

2. TylerGrunter

Posts: 1544; Member since: Feb 16, 2012

It was quite fishy the score in Antutu for the Lenovo K900 and I always said to wait for the results of some real CPU tests like Geekbenck before crowning the Z2580. It just turned out to be true. Kind of expected, wasn´t it?

18. tadaa

Posts: 267; Member since: Apr 18, 2013

So much for lenovo's marketing

3. Edmund

Posts: 656; Member since: Jul 13, 2012

The exynos chip consumes more power; It also has 6 more cores

5. iushnt

Posts: 3105; Member since: Feb 06, 2013

its not 6 more cores, its only 2 more cores when it comes to real time operational performance..even though it is said to be consuming more power, it is still the second best battery performer in its category after gs4 sd600 version..Its only slightly lower than gs4 sd600 but much higher than lumia, xperia z, iphone 5 and htc one

7. Shatter

Posts: 2036; Member since: May 29, 2013

At max power consumption the atom uses about half what an a15 quad core uses. Atom is the really crappy low end chip that intel makes. Wait till the other chip series get their power consumption low enough to use in phones. Baytrail will be siginficantly better than A15 but as of right now it looks like its $200-$400 range tablet only.

13. livyatan

Posts: 867; Member since: Jun 19, 2013

Can you be more wrong? Atom Z2560(1.6GHz) TDP - 6.7W A15 Exynos dual(1.7GHz) TDP -9.1W Atom E3850 (Bay Trail-I) - 1.91GHz quad core, TDP 10W, 792MHz GPU, price around 150$(estimate based on the $132 price of the Quad 1.60GHz Bay Trail-M Celeron with 7.5TDP and 756MHz GPU) Silvermont looks ridiculously overpriced and will be nowhere near competitive.. the ARM based SoC's will trash it it in GPU, be roughly on par with CPU, and far more affordable for the OEM's.

14. hung2900

Posts: 966; Member since: Mar 02, 2012

8W TDP of Exynos 5250 (as i remember from Anandtech, not 9.1W) is only theoriacally when both CPU and GPU run maximumly. However, this never is true due to throttling system (limit to 4W only) and no task needs that much power. Don't be wrong between TDP and real battery usage, and between Exynos 5250 and Exynos 5410

17. livyatan

Posts: 867; Member since: Jun 19, 2013

Go to notebookcheck.net (best reviews on the net imo). On the Nexus 10, they measured 9.2-9.4W. Interestingly enough, for the Atom Z2760 with a paltry GPU, they got an even higher value - 9.5W

15. sorcio46

Posts: 435; Member since: Jul 27, 2011

Where did you find the TDP values?

16. livyatan

Posts: 867; Member since: Jun 19, 2013

I cannot post links so I'll just give a copy paste breakdown. It's an official data from Intel. -Atom E3810 – 1.46 GHz single core CPU w/400 MHz GPU and 5W TDP -Atom E3821 – 1.33 GHz dual-core CPU w/533 MHz GPU and 6W TDP -Atom E3822 – 1.46 GHz dual-core CPU w/667 MHz GPU and 7W TDP -Atom E3823 – 1.75 GHz dual-core CPU w/792 MHz GPU and 8W TDP -Atom E3840 – 1.91 GHz quad-core CPU w/792 MHz GPU and 10W TDP Intel Bay Trail-M (Celeron, Pentium for notebooks, convertibles) -Celeron N2805 – 1.46 GHz dual-core CPU w/667 MHz GPU and 4.5W TDP -Celeron N2810 – 2 GHz dual-core CPU w/756 MHz GPU and 7.5W TDP -Celeron N2910 – 1.6 GHz quad-core CPU w/756 MHz GPU and 7.5W TDP -Pentium N3510 – 2 GHz quad-core CPU w/750 MHz GPU and 7.5W TDP Intel Bay Trail-D (Desktops) -Celeron J1750 – 2.41 GHz dual-core CPU w/792 MHz GPU and 10W TDP -Celeron J1850 – 2 GHz quad-core CPU w/792 MHz GPU and 10W TDP -Pentium J2850 – 2.41 GHz quad-core CPU w/792 MHz GPU and 10W TDP Theres also unoffocial mention of the supposed quad core 2.4GHz(??) mobile reference chip Bay Trail T, with 2.5WSDP and no TDP revealed but it should be at least in the bellow 5W range, to compete with Snapdragon 800.

22. Shatter

Posts: 2036; Member since: May 29, 2013

Baytrail has a ton of variants. Some of them are geared towards battery life others for maximum power.

25. TylerGrunter

Posts: 1544; Member since: Feb 16, 2012

That´s what Intel and Intel fans have been saying for the last 3 years: wait till XXXX comes out and will smoke out ARM chips. And for the last 3 years it never came true. Will it ever be true? I really start to wonder.

4. FlushGordon unregistered

There has been misleading reports about the K900's micro sd card support, does it have one?

6. Shatter

Posts: 2036; Member since: May 29, 2013

No it does not.

8. livyatan

Posts: 867; Member since: Jun 19, 2013

Well, well..I kept telling how memory scores for Atom in Antutu are very misleading and that the bench needs to be ignored..the fact is that Clover Trail + is beaten by 50-150% in other benchmarks. So yep, TOLD YOU SO, Intel fanboys! BTW, the graphics tests on Antutu are worthless also.. I noticed basically the same scores for the Galaxy S3 and Galaxy S4.

9. FlushGordon unregistered

I'd much rather trust 3DMark Some flagship models tend to deflate-inflate scores (depending on the AnTuTu version) The HTC DNA was a prime example, now we have the K900

11. Pedro0x

Posts: 271; Member since: Oct 19, 2012

Antutu is worthless, it is so easy to optimize for your phone. You can optimize Nexus 4 so it has 29000 points(I think) which is more than SGS4. And yes, atom do not have good memory scores, but Silvermont should change that.

23. Shatter

Posts: 2036; Member since: May 29, 2013

SIlvermont has DDR3 support, I think its a bit higher mhz than a15/krait can do too.

10. kanagadeepan

Posts: 1260; Member since: Jan 24, 2012

Another proof that benchmark scores have nothing to do with real life performance and hence purchase decisions should never be based on them, imho...

12. FlushGordon unregistered

Oh Come on! With your reasoning, you wont rather care if the phone you're purchasing is stuffed with a crappy Adreno 203 or a PoWERVR544mp Benchmarks pretty much give you the idea (altho not always precise) on how they'd perform in the real world

24. Shatter

Posts: 2036; Member since: May 29, 2013

Not really the mediatek turbo @1.5ghz runs everything smoothly and can play every 3d game on the Play Store. After a certian point the benefits from more power start declining. More power is not bad at all but it does use way more power. A phone with a mediatek turbo can be expected to get around 2 days of battery life atleast with regular use. A15 on regular use will require a charge almost every day.

26. XiphiasGladius

Posts: 813; Member since: Aug 21, 2011

I’m waiting for midrange phones (price wise) powered by that particular Mediatek chip to hit the mainstream market.

27. FrostNerdy

Posts: 1; Member since: Jul 14, 2013

I don't care about the Benchmark, it depends upon the user, just buy GS4 and at the same time Lenovo K900, then compare them! Check which is the best between the 2, sometimes Benchmarks are unrealiable.

Latest Stories

This copy is for your personal, non-commercial use only. You can order presentation-ready copies for distribution to your colleagues, clients or customers at https://www.parsintl.com/phonearena or use the Reprints & Permissions tool that appears at the bottom of each web page. Visit https://www.parsintl.com/ for samples and additional information.