INTRODUCTION
-------------------
First off, I have to mention that these benchmarks of the e450 were done on ASUS 1215B, there were done not the test the e450 itself but rather the Super Hybrid Engine that is installed with the 1215B. Although these results might be of interest to fellow 1215B users, I would be curious to find out if other e450 users, like the DM1, are experiencing similar things. I'll also add here that while I am quite a knowledgeable computer enthusiast, I'm not a computer engineer. And sorry, forgot to take screenshots. If it's it TL;DR for you, then just scroll down to the ANALYSIS and CONCLUSION sections to get the highlights.

To begin, I undertook these test mainly to measure certain speed deficiencies that I noticed while getting to know my laptop, which I received less than two weeks ago. I decided to do some tests to measure the e450 cpu frequency under various conditions. I'm also sharing these tests so that potential buyers can find out a bit more about the products (e450, 1215B), and hope that other users will share their results as well.

TESTING RESULTS
---------------------
To do the tests, I used the following software:

1. HWINFO64 / Speccy / CPU-Z: To measure the CPU frequency under load. Yes, I used all three software mainly because I was getting some weird readings that I could not understand at first.
2. Super PI 1.1: CPU load test, with a timed end result for comparison.
3. ASUS SHE or Super Hydrid Engine: To test the performance of the 1215B under 3 conditions: High Performance, Auto Mode / Auto High Performance, and Battery Saving.


--------------------
Test Series 1: The first series of tests (Super Pi 1MB) involved testing the 3 hydrid engine with the notebook plugged in and battery levels above 93%,and power plan set to max performance:

a) High Performance Mode / Battery at 95% with Adapter Plugged in / CPU max@1646MHz: 48s
b) Auto High Performance Mode / Battery at 95% with Adapter Plugged in / CPU max@1646MHz: 48s
c) Power Saving Mode / Battery at 95% with Adapter Plugged in / CPU max@1317.9 MHz: 60s

As can be evidently seen, CPU speeds are max out at 1646MHz in High Performance and Auto modes. Choosing a power saving mode even while your power plan is set to max performance will cause the CPU to max out at only 1317MHz.

--------------------
Test Series 2: The second series of tests (Super Pi 1MB) involved testing the 3 hydrid engine with the notebook unplugged and battery levels above 93%,and power plan set to max performance:

a) High Performance Mode / Battery at 95% with Adapter unplugged / CPU max@1646MHz: 50s
b) Auto High Performance Mode / Battery at 95% with Adapter unplugged / CPU max@1317MHz: 60s
c) Power Saving Mode / Battery at 95% with Adapter unplugged / CPU max@1317MHz: 60s

Unplugged, high performance mode still maxes the cpu to the full 1646Mhz, but Auto and Power Saving modes defaults down to 1317Mhz.

--------------------
Test Series 3: The third series of tests (Super Pi 1MB) involved testing the 3 hydrid engine with the notebook unplugged and battery levels below 90%,and power plan set to max performance:

a) High Performance Mode / Battery at 89% with Adapter unplugged / CPU max@1646MHz: 50s
b) Auto High Performance Mode / Battery at 89% with Adapter unplugged / CPU max@1317MHz: 59s
c) Power Saving Mode / Battery at 89% with Adapter unplugged / CPU max@1317MHz: 59s

So the third test pretty much shows that same results, that there are no difference when comparing battery levels of 95% and below 90%. So why this test? To contrast directly with the next and final test, done immediately after test series 3 but with the notebook plugged in.

--------------------
Test Series 4: The fourth and final series of tests (Super Pi 1MB) involved testing the 3 hydrid engine with the notebook plugged inand battery levels below 90%,and power plan set to max performance:

a) High Performance Mode / Battery at 88% with Adapter Plugged in / CPU max@823MHz: 90s
b) Auto High Performance Mode / Battery at 95% with Adapter Plugged in / CPU @823MHz: 90s
c) Power Saving Mode / Battery at 95% with Adapter Plugged in / CPU @823 MHz: 90s

What is clear here is that at battery levels below 90%(ish), CPU speed is limited to only 823MHz, even when the power plan is set to max performance AND power adapter plugged in. In this case, it does not matter what your hybrid engine setting is, or your power plan for that matter - max CPU speeds are limited until battery levels increase (in my case, to about 92%).


ANALYSIS
------------
My initial reasons for testing the e450 and the Super Hybrid engine was to make sure it works. And it does for the most part. Obviously the SHE is designed to be most effective when a user is on battery, and allows the user to choose to maximize battery or performance. But what really surprised me most is that SHE does not work when plugged in and battery levels are below 92%. Worse still, when plugged in and battery levels are below 92%, the CPU speed is WORSE compared to when it is in battery mode and SHE set to Power Saving Mode:

TEST 4) High Performance Mode / Battery at 88% with Adapter Plugged in / CPU max@823MHz: 90s
TEST 3) Power Saving Mode / Battery at 89% with Adapter unplugged / CPU max@1317MHz: 59s

It would appear that the culprit here is the small and cute but obviously underpowered AC adapter, which seems to be incapable of both charging the notebook battery and at the same time juice the CPU up to max speeds. Unfortunately, the issue of crippled CPU while charging is not unheard off, with many Macbook users complaining about the same issue. But realizing that I'll get crap performance from my CPU while plugged in until my battery is up to 92% -- it's a bit nuts. What is unfortunate in the case of 1215B is that customers have no option to purchase a more powerful adapter.

Thankfully, I can tell you that the 1215B runs at full performance mode when battery is removed completely and notebook running on AC power alone. This should be your setup when playing games on laptops anyway, as the heat is not good for the batteries in the long run. I also say "thankfully", because for certainly laptops such as some / most lenovos, running their laptops on ac power alone without the battery will result in a stepped down CPU performance (based on my research while considering the e120). According to Lenovo, it is a design feature. Personally, I think it is a design flaw.

CONCLUSION
---------------
Based on my personal test results it would be appear that:
1) For 1215B users, to get max performance from your machines, you should be plugged in with your battery above 92% (or take the battery out completely) and your SHE set to High Performance. Here, your CPU will max out at 1646MHz.
2) To get the worst performance, have your notebook plugged in while battery is below 90ish%, and your CPU will only max out at 823MHz

I would be interested to know if other 1215B users are getting similar results. If you guys are, then it would appear that it's a design issue with the 1215B. Having said that, I would really be interested to know if DM1 users are having the same results. Then, it would appear to be a e450 issue, possibly.... If no one else is having this issue...

Hope you found this useful.

This post has been edited by Isky: Dec 19 2011, 04:21 AM

1646MHz is the highest cpu speed for the e450 (1.65MHz). The e350 was rated at 1.60MHz. I believe the TurboCore refers to the GPU / Radeon part, which I did not measure.

Tested on Win7 Pro 64 SP1. No, the SHE does not overclock for the AMD apu, since as far as I know there is no way to overclock the e450, except on non-laptop based motherboards. I don't know about the intel series though. The SHE is more of a power-saving feature, and underclocks the apu to help extend your battery life.

Yup. This tool can only be used to underclock / undervolt, not overclock. As i mentioned, I think the only e450s that can be overclocked are pc motherboard-based, not laptop-based.

Yeah, finding it was a little surprising. But overall, much better than other laptops. The 1215B at least runs full speed when unplugged, or when charged above 93%. So it's really a matter of making sure u are fully charged or simply unplugged to get full speed. For me, I like to game without my battery in, so I still get my full speed. However, lenovo laptops are crippled if their batteries are not plugged in. I read a thread on their forum, where someone complained about the design flaw, but the engineers kept insisted it was a design FEATURE. I had been considering an x120e/x121e at the point, but after reading the thread, I struck lenovo completely off my life of brands to consider.

I assume that the speed is crippled while charging to ensure that the battery is charged as quickly as possible. However, I would have preferred it if Asus gave users a choice to either prioritize charging or speed, or give us the option to buy a more powerful charger.

To be honest, I'm not sure. I suspect the energy mgmt system is probably in the bios, otherwise one might have different results between Win7 and WinXP. At the core of it, I think (as I mentioned earlier) the biggest factor for the reduced speeds while charging is the less powerful adapter, so it's really a combination of bios and hardware. I remember a similar argument being made at the lenovo forums.

FYI, I reformatted my drive from win7 starter to win7pro. I still have the speed drop while charging. I kept the energy mgmt software since it will help to extend battery life.

This post has been edited by Isky: Nov 26 2011, 02:32 PM

Thanks for the update vase_t. I'll do a retest to see if the new rom does what is promises to do, but it quick check seems to verify that the problem has been fixed. Glad that this issue has been neatly resolved.

Harveylee, the e450 has its plus and minuses. Whether it is good for you depends on what you intend to use it for. It's better than the older atom processors, more than adequate for general word processing and surfing, while the Radeon HD 6320 lets you do some decent gaming on the side. But if you intend to do a lot of cpu-intensive work like photoshop, 3d animation, etc., the i3 or the i5 is the way to go. Of course, it's all about tradeoff - budget, speed, size and weight.