Overclock.net › Forums › Intel › Intel CPUs › [EETimes] Has Intel Really Beaten ARM?
New Posts  All Forums:Forum Nav:

[EETimes] Has Intel Really Beaten ARM? - Page 4

post #31 of 32
They've been doing this for years already, iirc it was originally using the family ID to detect if a CPU could run SSE3 (For example, you can replace SSE3 with virtually any instruction set) rather than actually just detecting normally.
What was the effect of Intel doing that? Naturally, their CPUs ran the latest code while AMDs were stuck with SSE2, SSE and x87, it's the same crap here.
post #32 of 32
Originally Posted by Heuchler View Post

AnTuTu newest version - Clover Trail drops 50% in memory performance, 20% cpu - ARM Versus Intel - Part 2

ARM Beats Intel With Revised AnTuTu Benchmark @ EETimes

Earlier this month, the AnTuTu benchmarking tool raised a bit of a firestorm when it showed Intel‘s Atom Z2580 out performing offerings from Qualcomm, NVIDIA, and Samsung. This led to BDTI, a consulting firm, showing that not all instructions were being executed by the Intel processor, causing a false boost in results. As such, AnTuTu has issued a revised version of the benchmarking tool, and Intel has fallen 20-percent as a result. [50% in memory test portion of AnTuTU for Clover Trail while Snapdragon 600 stayed the same].


The EE Times story was written by analyst Jim McGregor of Tirias Research, who is a credible analyst that we have personally met and known for years. He recently noticed some discrepancies in a new version of the AnTuTu benchmark that made him to a double take. For years we have seen companies cheat on benchmarks when it comes to computers, so are companies cheating or making sure benchmarks are enhanced for their technology? After some digging he found that the ICC compiler introduced around version 2.9.4 of the AnTuTu benchmark was used just for the Intel processors. It appears that AnTuTu was aware that they were 'busted' and issued revision 3.2.2 to the benchmark Wednesday evening. Mr. McGregor went back and benchmarked AnTutu v3.3.1 versus v3.3.2 and the differences are rather alarming and it was clear that the benchmark favored Intel processors more than other companies.

Legit Reviews

The release notes of the latest build of AnTuTu Benchmark show that changes were made that 'Enhances the stability of scores'.

To make matters worse, we ran across this post on Berkeley Design Technology's (BDTI) Web site found that "the ARM-based [Samsung] Exynos processor performs all the operations specified in the benchmark source code, while the Intel Z2580 processor skips some steps." Guess what benchmark this issue was found on? The AnTuTu Benchmark once again. Double Ouch!

We all know that cheating on benchmarks has been a big deal when it comes to computers and it appears that trend will continue on the mobile side. It looks like we need better smartphone benchmarks that aren't bought off and offer good honest real world tests!

PCWorld - Analyst questions tests that put Intel's 'Clover Trail' over ARM

Lenovo's K900 is a 5.5-inch 1920x1080 smartphone based on Intel's Clover Trail+ platform, with a dual-core Atom Z2580 processor clocked at 2GHz paired up with dual-core SGX 544MP2 graphics

thanks for the post, very interesting stuff.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › [EETimes] Has Intel Really Beaten ARM?