Nvidia, Asus Question Apple's iPad and Tegra 3 Comparison

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.
Also, has anyone else noticed the recent upheaval of Apple's many scams being brought to light? I'm not jsut talking about the ongoing Foxconn fiasco, but a lot of things are coming up like how Steve and his Apple participated in eBook price fixing.
 
[citation][nom]blazorthon[/nom]A5 has among the best graphics in any of the ARM devices, if not the best, so it;s not unreasonable for the A5X to be even three or four times more GRAPHICALLY capable than Tegra 3. However, it will not be computationally better. However again, that doesn't matter if the A5X's two cores are fast enough for it's graphics, which they undoubtedly are. Now it needs to have enough memory bandwidth to support the graphics, then there's the rest of the phone and it's connectivity, and more.Regardless of that, Apple should have provided hard facts and realistic benchmarks and none of the using non-optimal condition crap on the competition. Apple likes to jsut say they are better rather than prove it. This makes me assume that the A5X really isn't that much better than the A5 even though it probably is. Remember, just like in CPUs, core count is NOT everything in graphics. IPC, clock rate, memory bandwidth, and more are also important.Intel's dual core CPUs beat AMD's everything in gaming and Nvidia's cards have far fewer cores than AMD's but they more or less compete anyway. IE the GTX 580 has 512 cores, 7950 has 1792 cores, but they have similar performance at stock anyway.[/citation]

Explain how all benchmarks out there show the Tegra 3 having 10-20 more frames than the Ipad 2? Apple is making stuff up to look better. If anything the A5X is 2x more powerful than Tegra. Not 4.
 
[citation][nom]jefforange89[/nom]According to these benchmarks (http://www.anandtech.com/show/5163/asus-eee-pad-transformer-prime-nvidia-tegra-3-review/3), the A5 is 1.1-2x as fast as Tegra 3… so if the A5X is twice as powerful as the A5, then it would appear Apple's claims are accurate.[/citation]
That test was before ICS was available, which improves the performance. However, the Tegra 3 still isn;t producing any actual test results close to where it SHOULD be, which is faster than an A5.

However, Tegra 3 is the ONLY tablet/phone processor out there that could beat an A5 - the rest fall way behind. And Tegra 2 just plain blows.
 
[citation][nom]jasonpwns[/nom]Explain how all benchmarks out there show the Tegra 3 having 10-20 more frames than the Ipad 2? Apple is making stuff up to look better. If anything the A5X is 2x more powerful than Tegra. Not 4.[/citation]

I'm not explaining anything. I don't care too much about it right now so I didn't bother to look into Tegra 3. I simply stated that A5 does have some of the best graphics and if A5X is twice as good as the A5, well then it's possible. I've also constantly heard of people saying that Tegra 3 isn't too much better than most of what's out there.

Once again, I don't care. However, if I must explain away benchmarks, keep in mind that pretty much every comparison of the smart phone/tablet devices is biased. We always see these benchmarks where one device is running an older version of it's OS and/or program and the other has a fully updated version etc etc. I'm sure you know this, so don't hate me for saying that it's possible unless you can show me irrefutable poof that it's not possible. Why don't you do your own benchmark or find a benchmark that you can prove is unbiased, then we can see what's what.
 
[citation][nom]darkchazz[/nom]"The ASUS Transformer's GPU is only 12-cores..."More cores doesn't always mean better performance..Remember the 8-core gpu in the Tegra 2 ? yeah that one was completely smoked by the 4-core Mali gpu in the exynos SoC, and I hate to say this, but both were vastly outperformed by the 2-core SGX543 MP2 gpu in apple's A5...[/citation]


That's the point Nvidia is trying to make. More cores means nothing, and even less than that in the GPU space. The tegra 3 is powerful because of the CPU performance, it's basically 4 ARM 9 cores with higher clock rates (yes yes I know there are more differences but I'm taking real basic) and a 5th extra core thrown in there. Tablet games are almost all CPU dependent at this point because of the lack of shading and actual game making techniques involved in their development.
 
[citation][nom]icepick314[/nom]I'm more interested in how a tablet has more processing and graphics power than Xbox 360 or Playstation 3...[/citation]

Actually XBOX 360 is really old hardware at this point.... Console game programming is just extremely optimized for the hardware. I would not doubt that a tablet can be significantly faster than an xbox 360.
 
Has anyone considered if benchmarking here and/or the games have good multi-threading support? It's possible that the Apple chip is faster on a per-core basis which is usually the case when you have less cores (heat dissipation/power reasons) than than multicore processors. It certainly is misrepresenting the facts, but if most apps/games aren't multi-threaded it could be a defensible position....
 
BBC Top Gear showed how a Mustang was more fuel efficient than a Prius.
They Drove the Prius around a track as hard as they could, all the Mustang had to do was keep up with it. Because the Mustang engine is so much bigger, it didn't have to work as hard to maintain the same speed. So the prius actually used more fuel.

Any test can be made to get the results you want. Doesn't mean the results mean anything.
 
[citation][nom]willard[/nom]I'd be surprised if Apple were so dumb as to blatantly use false advertising. I'm guessing they used cherry picked test conditions to show the A5X as favorably as possible, and Tegra 3 as unfavorably as possible.[/citation]

Look back at Apples marketing. It was nothing but blatant false advertising. Why would they change that now.
 
I am not up on how the ipad works but if it has a separate cpu and gpu chip and the tegra is a unified chip I wouldnt be tooting my horn about performance if I were apple. Plus as many have said they upped the resolution for the ipad3 and that means if they didnt up the gpu too they will have LESS performance even though they added an additional 2 cores to the cpu. Plus the benchmark that was used to test the ipad2 vs the tegra3 was rendering in 720p and rendering offscreen... who actually trusts that as a benchmark?

http://www.anandtech.com/show/5163/asus-eee-pad-transformer-prime-nvidia-tegra-3-review/3
 
[citation][nom]DaddyW123[/nom]BBC Top Gear showed how a Mustang was more fuel efficient than a Prius. They Drove the Prius around a track as hard as they could, all the Mustang had to do was keep up with it. Because the Mustang engine is so much bigger, it didn't have to work as hard to maintain the same speed. So the prius actually used more fuel.Any test can be made to get the results you want. Doesn't mean the results mean anything.[/citation]
Actually it was a bmw M3 with like 440bhp they used... but yeah your main point was correct.
 
I don't get it is the A5x supposed to be a Quad-core or Dual Core? The previous article on Tom's said it's a Quad-core but this one says its a dual-core? Which is it? Tegra 3 (Quad core) vs A5X (Dual Core)?
 
[citation][nom]DaddyW123[/nom]BBC Top Gear showed how a Mustang was more fuel efficient than a Prius. They Drove the Prius around a track as hard as they could, all the Mustang had to do was keep up with it. Because the Mustang engine is so much bigger, it didn't have to work as hard to maintain the same speed. So the prius actually used more fuel.Any test can be made to get the results you want. Doesn't mean the results mean anything.[/citation]

Its was a BMW. And you are right test can be conducted to be in favor of any particular outcome. Lets put the benchmarks aside and look at real world performance for apps. Microsoft did this for the WP7. It beat nearly every OS/device that came up against for actual app launches and user tasks such as posting pictures and sending texts. While it did not beat every device at every test. I did perform surprisingly well again the other major competitors. Do some benchmarks timing things that people will actually do.
 
[citation][nom]zzz_b[/nom]Everybody talks about the resolution and graphics.I do not want to take anything away from Apple, the display must be very nice. The question is how will it be used? The HD movies are 1920x1080. Games are not there yet either.Nobody is saying one word about USB, microSD, HDMI,... connectivity. I think these are very important too.[/citation]
who uses usb or microSD or HDMI honestly? nobody, thats who.
but everything is in 2048 x 1536 it would be foolish to bother with things such as usb or microSD. instead, Apple prioritizes what is trully important
 
I am under the impression that Tegra 3 is a disappointment, its been out for 6-7 months and only in a few devices. TI OMAP claims they are 2X faster than Tegra not too long ago!
 
I am under the impression that Tegra 3 is a disappointment, its been out for 6-7 months and only in a few devices. TI OMAP claims they are 2X faster than Tegra not too long ago!
 
Thank you guys for the correction - it's been a while since I saw that episode and I knew if I took the time to look it up, my comment would have ended up on page 3 here. So I just chose Mustang.
 
I'm thinking it better be 4x faster to push a reasonable frame rate on that retina resolution.

30 fps on a retina display is 4 times faster than 30 fps on a non retina display because its 4 times the pixel s.
 
[citation][nom]willard[/nom]I'd be surprised if Apple were so dumb as to blatantly use false advertising. I'm guessing they used cherry picked test conditions to show the A5X as favorably as possible, and Tegra 3 as unfavorably as possible.[/citation]
Apple isn't know for being truthful....

I wouldn't be surprised if they took nVidia's worse scores and the iPad's best scores, completely ignored the fact that the scores were produced by different tests and assembled a BS graph based on the scores they chose. This was the approach Steve Jobs used (if someone else has a better product, falsify specs to make your product appear superior in every way)....
 
[citation][nom]ericburnby[/nom]That test was before ICS was available, which improves the performance. However, the Tegra 3 still isn;t producing any actual test results close to where it SHOULD be, which is faster than an A5.However, Tegra 3 is the ONLY tablet/phone processor out there that could beat an A5 - the rest fall way behind. And Tegra 2 just plain blows.[/citation]

if krait lives up to promises it should prove to be a tegra 3 killer
http://www.anandtech.com/show/5563/qualcomms-snapdragon-s4-krait-vs-nvidias-tegra-3
 
Honestly, I don't really understand synthetic benchmarks. Most of them i've seen, when it comes to graphics, kick out a score, that never holds water when it comes to actually playing games. I've seen many cards that benchmark higher in synthetics, but then loose badly to it's rival. Thats why real world benchmarks tell the real story. Same thing for 2d 3d performance, some of the new top of the line graphics cards have awesome 3d performance, but when it comes to 2d, they get there ass handed to them on a spoon by older graphics cards. Also, the software, drivers, make a difference too. Just because your hardware is 4 times faster, doesn't mean it's going to perform faster on the software. Optimizing makes a huge difference.
 
Gee what's apple going to say next, it can run crysis, which I'm sure apple will try to fake like the benchmarks. Apple probably rigged the benchmarks to favor the A5x or hacked the program to to show higher numbers. Heck I have a celeron d 356 with a windows experience index of 9, does that mean it's faster than an i7 no, does that mean I hacked it to look faster yes, does that mean apple done the same thing with the ipad yes.
 
Status
Not open for further replies.