Transformer Prime Benchmark

You should know by now just how we feel about benchmarks -- they're easy to manipulate, and really don't tell you anything about how a device will actually behave. And so you should know by now that we're really not going to get all that worked up over the ASUS Transformer Prime showing up in a benchmark app. Look, it'll be fast. It'll be great. Hell, we got to experience some early quad-core processor action some nine months go. It's going to rock. And that a quad-core device tops out like you see here should really surprise no one. But, anyway, there it is. Huzzah.

Source: Tech2; via Engadget

 
There are 35 comments

Reader comments

Shocker: Quad-core devices benchmark higher than single- or dual-core

35 Comments
Sort by Rating

I thought the GS II would have benchmarked better than the Galaxy Nexus. Nice to see it doesnt though :P since it will be my entrance into the Android experience from iOS.

I'm actually shocked by how similar the Galaxy Nexus and SGSII scores are. I thought the Nexus was supposed to be a performance improvement and not just ICS on top of the same old hardware capability?

Phil is no doubt reading these comments and mentally comparing writing for us to herding cats - we never stay on the topic at hand. LOL. Look, something shiny!

That's probably why the Nexus has the CPU clocked at 1.5 rather than 1.2 GHz. Also, I bet the GSII would still bench higher if it didn't have Touchwiz (but of course, benchmarks mean nothing, especially when the GSII has less pixels to push out).

I don't think there are any apps that take advantage of dual core or quad core processors are there? Its been a while but last I heard there were really none out there.

Applications often times spin up multiple threads to handle tasks. This is a common thing done in android to avoid Android Not Responding (ANR) messages. Developers will run intensive code on a separate thread to avoid locking the main GUI thread. If an app uses threading then it will definitely benefit from having multiple cores. Also, from what I understand ICS improves on the multi core support and or experience.

All apps coded with good foreground/background process style will benefit from dual core, but only on Android 3.0 and above. In other words, most smoothly-running apps already do their work in a background thread. Before Honeycomb, even on dual-core devices, these background threads were all run on the same core along with the UI and garbage collector for that app.

I'm not 100% certain, but I believe the only benefit you may get with <3.0 is some system processes or background apps may run on that other core. I'm not sure if that's built in (I know some vendors have made custom extensions) but the apps themselves are still limited in that case to one core at a time.

Now, though, any app that effectively splits work into separate threads could benefit. Basically, read this, from the Honeycomb release docs:

http://developer.android.com/sdk/android-3.0-highlights.html#multicore

And what is the point other than to get you to spend more money? I have a single core 800mhz phone that runs everything perfectly. I can even overclock it to over 1800mhz if I want but I have no reason.

You can:
- Impress your phone geek friends with your benchmark scores
- Use it as an excuse to trade up from your 2-month-old and now obsolete Android tablet
- Open you Gmail app twice as fast

Sorry for being off topic. My Asus Eee Pad Transformer should be arriving today via UPS. Did I make a wise investment, or should I have waited for something like the Prime?

Thanks

If you got it for $350 then I would say yes good choice. The Transformer Prime will sell for $500 when it comes out and it just provide a thinner, lighter body with a double the cores.

Sweet. Great to see the Nexus and the Galaxy SII so close in scores. Yes, I know these don't always translate into real world performance but it is encouraging.

I'm somewhat of a noob but uh... I think the most important "app" of them all supports multiple cores, Android. As for the 800 mhz phone running fine... Good, it's supposed to. That doesn't mean that more hardware dependant software won't be developed that will run better with a quad core. I doubt you're using a 10 year old computer for that very reason.

Just for fun... I'd love to see a benchmark comparison between my current OG Droid and my future new Galaxy Nexus! I'd like to try and hold out...but, not sure I can wait 3-6 more months for a quad-core in a handset though!

Note that the only things that show a substantial improvement are the RAM, the floating point variable calculations and the integer calculations. I'm sure the quad core experience will be absolutely amazing, but the lack of an improved graphics score does have me questioning how it will hold up against something like the Galaxy Nexus. I'd be willing to bet that this is a good case where the benchmark isn't telling the whole story.

Ah. You're obviously smarter than I am. I said the same thing, but you know what you're talking about. These benchmarks just clinched my decision to purchase the Galaxy Nexus. It doesn't mean a lot in the real world, but it's certainly the icing on the cake.

Keep some perspective.

This is one benchmark, run on a tablet (and a phone) that hasn't even been released yet.

Not saying it is incorrect, but to think that it is deffinitive is a diservice to yourself.

-Suntan

If you look closely, it doesn't seem to perform much better with 3D graphics than any other dual core device. Is that because apps/software haven't yet caught up with quad cores? The only massive improvement is with RAM, CPU Integer, and CPU float point. I'm not techy enough to understand what the last two mean, but to me the 3D graphics score show us that quad core are largely uneccessary at this point. For goodness sake, my 3 year old computer only has a triple core processor. Sheesh.

This is not entirely true. Obviously the scores for 2D and 3D graphics show up to be similar to something like the Galaxy Nexus, but I'm guessing that those are scores based off of the performance of code running in the OpenGL environment. A game most definitely could have code and calculations being performed outside of the GPU as well which would definitely improve the gaming experience. Also, the RAM score is substantially higher which should definitely improve on the gaming experience. Faster RAM will almost always yield a noticeable performance improvement.

For anyone interested in gaming a benchmark showing low, average and highest FPS would definitely be more telling than these numbers we see here. This image is only a small piece to the puzzle.

It's been stated multiple times that the graphics benchmark hit a capped limit and is not properly testing the GPU of the Tegra 3.

The question I would ask is not whether a quad is useful now but whether it will be useful in two years.

I would agree that most consumers will find them useless, Just not all. Most consumers also don't need an i7 but they still sell well.

Few people buying quad core tablets (or phones) will keep them for 2 years. I see the same argument for buying $600 video cards - that they will be usable longer. But the fact is that first adopters trade up constantly so longevity is irrelevant.

What I am seeing here is the galaxy s2 running gingerbread putting out similar numbers as a hrdware optimized galaxy nexus running ics. Hmmmm am I understanding this correctly?

Okay for all those wondering why the galaxy nexus and the SGS II have similar scores is because the SGS II has a screen resolution of 800x480, while the galaxy nexus has a resolution of 1280x720, which means that a difference of 480x240 pixels, and if it takes the SGS II to process what it does and get that benchmark score, just think about how much more processing power it takes the galaxy nexus to get the same score with a much larger amount of pixels to use. A similar but less extreme example is between the Motorola Atrix and the Optimus 2x. The while the Optimus 2x score higher than the Atrix, the Atrix has to make use of a larger 960x540 resolution instead an 800x480. If the Atrix and Optimus 2x had the same resolution, they'd have an almost identical benchmark score. And if the galaxy nexus had the same resolution as the SGS II, and the benchmark for the galaxy nexus would be far supioror to the SGS II's