iPad 2: GPU Benchmarked with Impressive Gains

[Lesser Evets]

This and other benchmarks make me very hopeful for the iPad becoming a powerful home-portable in a couple years.

I took out my old orange iBook 300, and it's incredible how 10 years later Apple made a huge leap in speed and compacting computing along with extended battery life. Can't wait to see what comes out in 10 more years.

Don't worry, in a few months, they will meet or exceed the abilities of the new iPad.

In about 13 months?
Will they ever meet the price and the hardware and the overall product?

 

[JordanNZ]

So,

1) the iPad only has a 900MHz dual core instead of the advertised 1GHz
2) the CPU is only about 50% faster and not up to 2x
3) the GPU is only about 4-7x faster and not up to 9x

Ouch.

You can't be serious with this post.....

 

[TomCondon]

Wirelessly posted (Mozilla/5.0 (iPod; U; CPU iPhone OS 4_1 like Mac OS X; en-us) AppleWebKit/532.9 (KHTML, like Gecko) Version/4.0.5 Mobile/8B117 Safari/6531.22.7)

You know, all the screen problems, etc and now they have to come up with this half baked benchmark to show how good a crappy product is, wew!

Get a grip, the iPad has seen its day, time to move on.

Who wants a third peice of electronic junk to carry around anyway?

Not I.

You will seriously eat those words about the iPad being "done" lmao


No one forces you to buy one, so why do you care if others do?

 

[TomCondon]

Wirelessly posted (Mozilla/5.0 (iPod; U; CPU iPhone OS 4_1 like Mac OS X; en-us) AppleWebKit/532.9 (KHTML, like Gecko) Version/4.0.5 Mobile/8B117 Safari/6531.22.7)

Apple fanboys have said that specs don't matter. Now they do? It seems to me they only say this when they know it should be more.

You're confusing specs with performance measurements. The former are just numbers, the latter directly affects the user experience (tm), the optimisation of which is Apple's primary focus.

This article shows performance test results. You are right that the article does talk about some specs such as numbers of pipelines and chip numbers etc, but that is not the main point of the article, nor has a single post in this thread to date mentioned those specs. The main point are the test results, which directly improve user experience (tm).

Well No you see, people here do care about specs, but believe user experience and design and build quality is paramount.

Sure the 13" MVP could have a dGPU, but then how would it have such good battery?

The iPad just happens to have both form and higher specs

 

[zacman]

You can't be serious with this post.....

Why not? Apple didn't say "The iPad2 runs circles around the XOOM" but its CPU is up to 2x faster and the GPU up to 9x faster than on the iPad1. Typical benchmarks now show that this is not true for maybe 99% of the time. So for an iPad1 user this is much more interesting than the stuipid comparision with the XOOM that nobody buys anyway.

 

[JordanNZ]

So,

1) the iPad only has a 900MHz dual core instead of the advertised 1GHz
2) the CPU is only about 50% faster and not up to 2x
3) the GPU is only about 4-7x faster and not up to 9x

Ouch.

Let me try again.

The CPU speed scales up to 1GHZ, it will downclock itself dynamically if not much is happening to save battery life.

There are 2 cores on the CPU, when software is written to make use of both of them, it will run twice as fast (or faster) than the first iPad.

And UP TO 9 times faster is up to 9 times faster.

 

[newfoundglory]

Right, i'm definitely holding out for an iPhone 5 now!!! Bring on the A5!

 

[zacman]

Let me try again.

The CPU speed scales up to 1GHZ, it will downclock itself dynamically if not much is happening to save battery life.

Wait, wait. What you are saying is:

You run a heavily CPU centric program on the iPad (like a eh... CPU benchmark) and the iPad itself chooses not to use the full power of the CPU to run these programs. I call that a bug and not a "battery saving feature".

If a program needs the full CPU power it should get it no matter of battery life. Period.

 

[JordanNZ]

Wait, wait. What you are saying is:

You run a heavily CPU centric program on the iPad (like a eh... CPU benchmark) and the iPad itself chooses not to use the full power of the CPU to run these programs. I call that a bug and not a "battery saving feature".

If a program needs the full CPU power it should get it no matter of battery life. Period.

The benchmark didn't show the CPU frequency.

 

[zacman]

The benchmark didn't show the CPU frequency.

"Geekbench reports clock speed at 900MHz." Are you saying that this is wrong?

http://www.anandtech.com/show/4215/apple-ipad-2-benchmarked-dualcore-cortex-a9-powervr-sgx-543mp2/2

 

[JordanNZ]

No..

"Its 1GHz. Geekbench reports the instantaneous speed, so you'll hear different numbers from that depending on what it ramps its speed down to to save power."

Give it a rest...

 

[zacman]

No..

"Its 1GHz. Geekbench reports the instantaneous speed, so you'll hear different numbers from that depending on what it ramps its speed down to to save power."

Give it a rest...

So again: You're saying that it clocks down the CPU to 900Mhz even when you use CPU heavy apps. That's a bug and not a feature.

 

[JordanNZ]

So again: You're saying that it clocks down the CPU to 900Mhz even when you use CPU heavy apps. That's a bug and not a feature.

No.

They're saying that it's on it's way up to 1GHZ but hasn't reached that point yet when the clockspeed is reported...
Get it?

 

[daneoni]

I'm guessing the upped the GPU specs in preparation for the retina display and not solely for gaming?

 

[ten-oak-druid]

Poor xoom

 

[INeedAMacBook]

Funny, not only was everyone talking about the new iPad 2 then, but supplies of the iPad 1 were pretty much drying up. You mustn't do much research before a fairly large purchase

It's weird though that it's feeling outdated. Didn't you just buy it 15 days ago? How is it no longer doing what you bought it to do?

And you could've returned it with in the 14 days when the invitation came out (which was so obviously for the ipad 2 you'd need an extra chromosome to not figure it out).

 

[ten-oak-druid]

Don't worry, in a few months, they will meet or exceed the abilities of the new iPad.

Cross your fingers and maybe your dream will come true.

LOL

Get some help.

You know some people get a tablet to enjoy viewing photos and movies on it. Others buy one just for the technical specifications and write "ipad killer" across the back with a silver sharpie.

 

[newfoundglory]

By the time any competitor manages to ship something which meets iPad specs (at a decent price too), we'll already be on iPad 3 with hi-res screen and the A6.

And the cycle continues

 

[fabianjj]

No.

They're saying that it's on it's way up to 1GHZ but hasn't reached that point yet when the clockspeed is reported...
Get it?

Well Apple isn't saying it's 1.0 GHz, anything in the range 0.5-1.49 GHz could still be considered 1 GHz in the mathmatical sense.

 

[*LTD*]

Don't worry, in a few months, they will meet or exceed the abilities of the new iPad.

Raw specs? Sure. Like a Dell running Winblows.

User Experience will be impossible to match. Horizontal vs. Vertical business models. Vertical done right will always trump anything else. Always.

Same as with the iPhone. Lots of Android devices, still no single "iPhone-Killer." The iPhone remains the best-selling handset on the market. And it's not due to specs.

Don't get caught in the specs trap like a lot of others around here.

What's REALLY funny is that Apple's incremental effort with the iPad 2 *still* trumped the absolute best the competition could manage. THAT is how utterly ******* clueless the also-rans are. But it's the same story with the iPhone. The competition needs to flood the market with a lot of cheap junk in order to achieve anything against it. And Apple is still the one making all the money.

 

[ryu3000]

Cross your fingers and maybe your dream will come true.

LOL

Get some help.

You know some people get a tablet to enjoy viewing photos and movies on it. Others buy one just for the technical specifications and write "ipad killer" across the back with a silver sharpie.

The only iPad killer is the iPad 2! It just killed the original iPad!

 

[marksman]

I am still waiting for someone else to match the experience of the original iPad, let alone the freaking iPad 2.

They have had over 9 months, since everyone knew a tablet was coming long before that...

I would love to see the 50+ scrapped tablets that have been burned by competitors of Apple in the last 12+ months because they were so poorly competitive.

 

[GnillGnoll]

1.) According to http://developer.motorola.com/products/xoom/ the Xoom has 27 native Open GL extensions. However, if you look at the benchmark, you see that it only displays 21 extensions, so the benchmark utility is not registering the additional extensions.

Motorola is listing extensions for OpenGL ES 1.1 here. GLBenchmark 2.0 is using OpenGL ES 2.0.

OpenGL ES 2.0 is not just an extended version of ES 1.1, it's a separate API and not backwards compatible. Each comes with its own list of extensions. The two lists actually have to be different.

2.) According to the benchmark aswell, it is registering the Xoom tablet as 1280x752.

Which is the actual resolution the benchmark is rendering at. You do realize that Honeycomb apps cannot use the entire screen? The System Bar at the bottom (apparently 48 pixels in height) is always there, though there is a "lights out mode" to dim it where it would be distracting (e.g. video viewing)

Which, by the way, means that the Xoom is really only rendering 22.4% more pixels.

"Geekbench reports clock speed at 900MHz." Are you saying that this is wrong?

The question is when Geekbench reads the clock speed. Is it during or before/after the actual tests are running? Clearly any mobile SoC should not run at its peak clock speed while it isn't heavily loaded.

 

[Kissaragi]

You know, all the screen problems, etc and now they have to come up with this half baked benchmark to show how good a crappy product is, wew!

Get a grip, the iPad has seen its day, time to move on.

Who wants a third peice of electronic junk to carry around anyway?

Not I.

That must be why they are hardly selling any of them.....

 

[CrArC]

Wow, those are some impressive numbers! Disclosure: I am an Android user, so I've watched this Xoom malarky with interest.

I checked out the rest of the benchmarks too. I'm hoping we can call them a fair representation of raw performance, in which case we can assume the maturity of the iOS platform is what lent it the upper hand since the processors themselves are identical between the Xoom and iPad 2.

Doubly interesting on the graphics - That NVIDIA seems bested at their bread and butter was a surprise. Or perhaps not; Imagination Technologies have far more experience in the low power market, after all.

It's going to get very interesting later this year and into the next, as NVIDIA are planning on their Kal-El revision (quad-core, twelve GPU cores) bringing as much as 5x the performance of the Tegra 2. I think there are videos floating about of this new tablet design running Xbox 360 games (only think - at least, they were intensive). The imaginatively named "Wayne" in 2012 should be an order of magnitude faster than Tegra 2. They're certainly not sitting on their laurels. By then we're looking at tablets which are more powerful than many desktop systems.

Love a bit of healthy competition! I think Apple and Android fans alike can look forward to cheaper, faster, and even more feature-packed tablets in the coming years.