DED must be nervous about the new Nexus line considering all the benchmarking articles he's been writing! It's interesting how benchmarks are suddenly very important.
Don't worry DED. The iPad will outsell the Nexus handily even if Lollipop provides a better user experience.
You must be on crack claiming that any Android OS is going to provide a better experience than iOS.
The Nvidia Shield isn't suppose to be a general purpose machine. It's a portable gaming machine meant for the gamer crowd.
But GPU technology is becoming general purpose, which means the need for niche hardware will get less and less as even low end general devices become very good at any application. There will only be faster and slower GPUs, not niche market GPUs. And the company that has the marketshare and technological lead in general GPUs will have no trouble selling lower end parts.
The one unusual part of this market is that no matter how good Apple's GPUs become, they will only go into Apple products, so nVidia doesn't necessarily need to keep up with Apple, but they do need to stay ahead of everyone else.
The Nvidia Shield isn't suppose to be a general purpose machine. It's a portable gaming machine meant for the gamer crowd.
They kept the resolution at 1080p for the same reason that PC gamers keep their resolutions at 1080p. Because that's the ideal resolution that graphics chips/cards that process complicated instructions - in this case Nvidia Shield wants to run half life 2 or portal at high FPS. Don't think they are trying for angry birds here.
You could flip this around too. There are intensive games like Bioshock that do NOT run on Apple's native resolution, as it's too graphics intensive. Games like that would run well on the Shield Tablet. I understand you're trying to make this an indictment of Nvidia's chip, but it's misleading to use this tablet as an example.
There's nothing "misleading" at all, because the article makes it very clear why the Shield uses a low / smartphone resolution, and doesn't ever suggest that it was "supposed to be a general purpose machine."
The problem for Nvidia is that the market for gaming-only tablets is tiny, and its vaunted Tegra K1 with what was supposed to be industry leading graphics is actually slower than Apple's general purpose A8X.
If you want to straighten the world out, why don't you leave comments on all the blogs who actually got this wrong and posted materially misleading hopes and expectations about Tegra (it's happened with every generation)?
Ok I have to say this. I hope Apple continue to expand into the GPU area and come up with hardware for Macs too at some point. How many of the MBP failure issues have been GPU issues? I rest my case.
Rumor has it that Apple has hired many of AMDs GPU designers so what you are wishing for inst impossible.
However do realize that GPU reliability issue has a lot to do with the competition between AMD and NVidia for the performance crown. They really haven't tuned their development to target reliability. GPU cards have been for years throw away items that get used for a couple of years and then upgraded to the latest and greatest. This has caused both AMD and NVidia to release hardware that is really operating beyond what would be ideal for longevity.
I suspect this is why Apple teamed up the way it did with AMD to produce GPU cards for the new Mac Pro. By allocating unique Apple part numbers they can tune the chips for reliability over performance. This seems to be the case because it appears that the Mac Pro uses AMD chips that are clocked lower than their mainstream "performance" cards.
Imagine an Apple GPU on a Mac that is faster than anything out there ... gamers would wet their pants. Well of course a lot of them do anyway but you know what I mean.
Well yeah but Apple still doesn't go after the GPU performance crown. They have always shipped conservative implementations that are reliable.
This is what I find interesting about the A8X, the implementation appears to be at the same time conservative but very high performance. A8X is a very very low power SoC yet it is well ahead of the rest of the world performance wise considering its power profile. This is what amazes me about the iPad Air 2. I'm shocked that people are so out of touch that they dismiss iPad Air 2 as a minor upgrade. It is in fact a fantastic update all around.
DED must be nervous about the new Nexus line considering all the benchmarking articles he's been writing! It's interesting how benchmarks are suddenly very important.
Don't worry DED. The iPad will outsell the Nexus handily even if Lollipop provides a better user experience.
So as an Android fan, you know that iOS devices have nothing to fear from Google's "Nexus" line (as Google itself says in its "we're not trying to introduce an iPad killer, just show Android partners how to build a product" PR), but you somehow think that the only writer who has been right about Nexus being a flop over the past 6 years is "nervous" about Nexus.
Apple probably already has A9X lined up for next year and A10X for the following year. It's scary.
That should be obvious! This chip is already good enough for a laptop and if they can extend the performance a bit with a clock rate bump could power an Air like laptop no problem. If they can deliver 14 nm chips next year we could see another massive increase in performance.
The K1 has decent performance but I suspect it is being held back by Google's interpreted/JIT code. No matter how much Google tries, native compiled code will always run rings around it.
These benchmarks are usually compile dwith native development kit (NDK), not with the interprted/JIT code.
But actually K1 is quite close to A8X, considering that Lollipop (Android 5.0) is still a 32-bit OS running on a 64-bit capable hardware. And I don't think we will see a 64-bit Android until next year, when K1 will be replaced by something else.
So as an Android fan, you know that iOS devices have nothing to fear from Google's "Nexus" line (as Google itself says in its "we're not trying to introduce an iPad killer, just show Android partners how to build a product" PR), but you somehow think that the only writer who has been right about Nexus being a flop over the past 6 years is "nervous" about Nexus.
In what way have Nexus devices been a flop if the intent is to introduce a new version of Android along with a base hardware platform for it. I think you yourself have noted they aren't meant to compete with licensees nor to be commercial successes, so how do they "flop"?
These benchmarks are usually compile dwith native development kit (NDK), not with the interprted/JIT code.
Yup. Most of the CPU and GPU benchmarks are natively compiled. The web-browser tests have to pass through a HTML and Javascript interpreter, so those tests are really CPU+JIT tests.
But actually K1 is quite close to A8X, considering that Lollipop (Android 5.0) is still a 32-bit OS running on a 64-bit capable hardware. And I don't think we will see a 64-bit Android until next year, when K1 will be replaced by something else.
Huh? 5.0 is 32-bit only? (I don't follow Android closely). And a 64-bit release is waiting on a 5.0.x or 5.x release in 2015?
For GPU, 32-bit vs 64-bit won't make much difference. Then, with Denver, I kind of wonder what the performance will really be like with 64-bit instructions. The code morpher, profiler looks to unroll loops and pack instructions into Denver's native VLIW format. So, is that VLIW instruction going to double in size for 64-bit to maintain the same number of packed instructions, or will the number of instructions that can be packed decreased by half for 64-bit instructions?
Is this really a shocker anymore? If there is a threshold for Apple's A-series chip designs we've yet to see it and I expect this gap to increase even further for a given power envelope for many years to come.
DED must be nervous about the new Nexus line considering all the benchmarking articles he's been writing! It's interesting how benchmarks are suddenly very important.
As it's always been, benchmarks in and of themselves have never been important when comparing disparate platforms, but the relevance is that the company that has always pushed ahead with better performance and usability due to unity of HW and SW is now also raw HW performance even with a lower power envelope. That's important because it shows that Apple's efforts to build their own chips — something that was deemed foolish when Apple bought PA Semi — is increasing the gap even further. If they can't been Apple in raw benchmarks there is simply no catch when using an off the shelf HW and SW compared to Apple's bepsoke designs.
In what way have Nexus devices been a flop if the intent is to introduce a new version of Android along with a base hardware platform for it. I think you yourself have noted they aren't meant to compete with licensees nor to be commercial successes, so how do they "flop"?
That all depends on if you believe's Google's inference that they don't want the Nexus line of products to be a market success. I think that's BS, but I wouldn't call myself partisan on this issue as I recently said something similar about Cook's comments regarding the hiding of the ?Watch in the accessories category.
Wha?? General purpose?, Like expect it in assorted Apple products like iPods and AppleTV's in addition to Apple's best tablet ever the Air2?
I had the same reaction, but I think he means the iPads are general purpose device, as opposed to other companies that will focus on some aspect of HW because they are focusing on a specific user. That said, I'm not sure I know of any modern tablets that's general purpose, unless you count an eInk-based device as a tablet that focuses on reading books.
In what way have Nexus devices been a flop if the intent is to introduce a new version of Android along with a base hardware platform for it. I think you yourself have noted they aren't meant to compete with licensees nor to be commercial successes, so how do they "flop"?
If a drunk falls down and then says he meant to fall down because he was too far from the ground to start with, is it not a fall?
But two things: first off, Android Enthusiasts predicted strong Nexus sales and continued to think that Nexus devices were not flops after many generation of Nexus flops. I've interviewed lots of Nexus owners, and virtually all of them think that the products sold well.
But now that it is getting impossible to say that Nexus launches have ever been anything other that wildly underachieving flops, Android Enthusiasts are backtracking to say that nothing Google does is ever a flop because the company is just spending billions of dollars trying to show third rate, incompetent hardware makers how they should be making great Android products.
It's like of like communists who say that communism isn't a flop, it's just a wonderful idea that has never been implemented correctly. How many years of failed attempts does it take to prove something is a failure?
Also, if you fact check yourself against reality, you'll find that ever since the first Google Nexus One, the company positioned Nexus as being "a new way to buy a phone," not a way to fail commercially or to not compete with licensees. Also, was Motorola also not "meant to compete with licensees nor to be commercial success"? Because, come on, that's just too much to swallow.
Now turn around and consider how vocal the same Android Enthusiasts were about Apple Maps and iPhone 5c being "flops," even though both have been wildly successful. Just ask Google, which has lost most of its iOS Maps user base and then lost a couple billion dollars trying to compete with Apple's low end phone with its own, unprofitable Moto X flop.
Really, your interlocking circles of logical fallacies are such incredible bullshit that it's hard to even comprehend how you can spew this stuff without hiding behind a phony identity. Oh wait never mind.
You must be on crack claiming that any Android OS is going to provide a better experience than iOS.
I already think it provides a better user experience thanks to the structure of the OS, but no 4.x build is as polished as iOS. I think 5.0 will match the polish of iOS and widen the gap in the sensibility and usefulness that Android already holds over iOS.
So as an Android fan, you know that iOS devices have nothing to fear from Google's "Nexus" line (as Google itself says in its "we're not trying to introduce an iPad killer, just show Android partners how to build a product" PR), but you somehow think that the only writer who has been right about Nexus being a flop over the past 6 years is "nervous" about Nexus.
You seem to be, but I'm not able to read your mind. Maybe you're just nervous about the software that they bring in tow. Lollipop appears to be poised to finally put Android on equal footing with iOS as far as polish goes. Considering the fact that a large number of iOS users attribute their choice to use iPhones to the polish and predictable performance of the software, I would imagine you might be nervous that iOS's one untouchable distinction might vanish. Thus, without any leg to stand on when criticizing the OS (the lates version anyway), you turn to the devices themselves and compare them in performance benchmarks, which you've always criticized as being unrelated to real usage.
I already think it provides a better user experience thanks to the structure of the OS, but no 4.x build is as polished as iOS. I think 5.0 will match the polish of iOS
Really? What Continuity and Handoff features has Google built for Mac OS X, Windows and Linux desktop so that you can seamless switch between devices?
and widen the gap in the sensibility and usefulness that Android already holds over iOS.
HA! Seriously though, you'll have show me what you mean by this Android sensibility that exceeds iOS.
I had the same reaction, but I think he means the iPads are general purpose device, as opposed to other companies that will focus on some aspect of HW because they are focusing on a specific user. That said, I'm not sure I know of any modern tablets that's general purpose, unless you count an eInk-based device as a tablet that focuses on reading books.
The context: "The problem for Nvidia is that the market for gaming-only tablets is tiny, and its vaunted Tegra K1 with what was supposed to be industry leading graphics is actually slower than Apple's general purpose A8X" indicates that what I mean was that Tegra K1 was supposed to be industry leading graphics, while Apple's A8X is not a dedicated gaming device or an extraordinarily expensive, special purpose niche device. It's going to sell in the tens of millions of units, running everything from enterprise apps to specialized imaging apps to videos games to Facebook.
Quote:
Originally Posted by wakefinance
I already think it provides a better user experience thanks to the structure of the OS, but no 4.x build is as polished as iOS. I think 5.0 will match the polish of iOS and widen the gap in the sensibility and usefulness that Android already holds over iOS.
sensibility "the ability to appreciate and respond to complex emotional or aesthetic influences; sensitivity. (sensibilities) a person's delicate sensitivity that makes them readily offended or shocked"
Yeah maybe that was the right word for "Lollypop" and its new web page appearance.
Comments
DED must be nervous about the new Nexus line considering all the benchmarking articles he's been writing! It's interesting how benchmarks are suddenly very important.
Don't worry DED. The iPad will outsell the Nexus handily even if Lollipop provides a better user experience.
You must be on crack claiming that any Android OS is going to provide a better experience than iOS.
The Nvidia Shield isn't suppose to be a general purpose machine. It's a portable gaming machine meant for the gamer crowd.
But GPU technology is becoming general purpose, which means the need for niche hardware will get less and less as even low end general devices become very good at any application. There will only be faster and slower GPUs, not niche market GPUs. And the company that has the marketshare and technological lead in general GPUs will have no trouble selling lower end parts.
The one unusual part of this market is that no matter how good Apple's GPUs become, they will only go into Apple products, so nVidia doesn't necessarily need to keep up with Apple, but they do need to stay ahead of everyone else.
The Nvidia Shield isn't suppose to be a general purpose machine. It's a portable gaming machine meant for the gamer crowd.
They kept the resolution at 1080p for the same reason that PC gamers keep their resolutions at 1080p. Because that's the ideal resolution that graphics chips/cards that process complicated instructions - in this case Nvidia Shield wants to run half life 2 or portal at high FPS. Don't think they are trying for angry birds here.
You could flip this around too. There are intensive games like Bioshock that do NOT run on Apple's native resolution, as it's too graphics intensive. Games like that would run well on the Shield Tablet. I understand you're trying to make this an indictment of Nvidia's chip, but it's misleading to use this tablet as an example.
There's nothing "misleading" at all, because the article makes it very clear why the Shield uses a low / smartphone resolution, and doesn't ever suggest that it was "supposed to be a general purpose machine."
The problem for Nvidia is that the market for gaming-only tablets is tiny, and its vaunted Tegra K1 with what was supposed to be industry leading graphics is actually slower than Apple's general purpose A8X.
If you want to straighten the world out, why don't you leave comments on all the blogs who actually got this wrong and posted materially misleading hopes and expectations about Tegra (it's happened with every generation)?
However do realize that GPU reliability issue has a lot to do with the competition between AMD and NVidia for the performance crown. They really haven't tuned their development to target reliability. GPU cards have been for years throw away items that get used for a couple of years and then upgraded to the latest and greatest. This has caused both AMD and NVidia to release hardware that is really operating beyond what would be ideal for longevity.
I suspect this is why Apple teamed up the way it did with AMD to produce GPU cards for the new Mac Pro. By allocating unique Apple part numbers they can tune the chips for reliability over performance. This seems to be the case because it appears that the Mac Pro uses AMD chips that are clocked lower than their mainstream "performance" cards.
Well yeah but Apple still doesn't go after the GPU performance crown. They have always shipped conservative implementations that are reliable.
This is what I find interesting about the A8X, the implementation appears to be at the same time conservative but very high performance. A8X is a very very low power SoC yet it is well ahead of the rest of the world performance wise considering its power profile. This is what amazes me about the iPad Air 2. I'm shocked that people are so out of touch that they dismiss iPad Air 2 as a minor upgrade. It is in fact a fantastic update all around.
DED must be nervous about the new Nexus line considering all the benchmarking articles he's been writing! It's interesting how benchmarks are suddenly very important.
Don't worry DED. The iPad will outsell the Nexus handily even if Lollipop provides a better user experience.
So as an Android fan, you know that iOS devices have nothing to fear from Google's "Nexus" line (as Google itself says in its "we're not trying to introduce an iPad killer, just show Android partners how to build a product" PR), but you somehow think that the only writer who has been right about Nexus being a flop over the past 6 years is "nervous" about Nexus.
#flawgic
That should be obvious! This chip is already good enough for a laptop and if they can extend the performance a bit with a clock rate bump could power an Air like laptop no problem. If they can deliver 14 nm chips next year we could see another massive increase in performance.
These benchmarks are usually compile dwith native development kit (NDK), not with the interprted/JIT code.
But actually K1 is quite close to A8X, considering that Lollipop (Android 5.0) is still a 32-bit OS running on a 64-bit capable hardware. And I don't think we will see a 64-bit Android until next year, when K1 will be replaced by something else.
These benchmarks are usually compile dwith native development kit (NDK), not with the interprted/JIT code.
Yup. Most of the CPU and GPU benchmarks are natively compiled. The web-browser tests have to pass through a HTML and Javascript interpreter, so those tests are really CPU+JIT tests.
Huh? 5.0 is 32-bit only? (I don't follow Android closely). And a 64-bit release is waiting on a 5.0.x or 5.x release in 2015?
For GPU, 32-bit vs 64-bit won't make much difference. Then, with Denver, I kind of wonder what the performance will really be like with 64-bit instructions. The code morpher, profiler looks to unroll loops and pack instructions into Denver's native VLIW format. So, is that VLIW instruction going to double in size for 64-bit to maintain the same number of packed instructions, or will the number of instructions that can be packed decreased by half for 64-bit instructions?
Apple’s iPhone 6 uses a 1334x750, 326-ppi pixel display
Shield tablet's 8-inch screen is 1,920 x 1,200, 283-ppi
iPad Air 2 screen is 9.7-inches, 2048×1536 264 ppi
Uhmm, isn't the Shield's resolution actually better?
Wha?? General purpose?, Like expect it in assorted Apple products like iPods and AppleTV's in addition to Apple's best tablet ever the Air2?
As it's always been, benchmarks in and of themselves have never been important when comparing disparate platforms, but the relevance is that the company that has always pushed ahead with better performance and usability due to unity of HW and SW is now also raw HW performance even with a lower power envelope. That's important because it shows that Apple's efforts to build their own chips — something that was deemed foolish when Apple bought PA Semi — is increasing the gap even further. If they can't been Apple in raw benchmarks there is simply no catch when using an off the shelf HW and SW compared to Apple's bepsoke designs.
That all depends on if you believe's Google's inference that they don't want the Nexus line of products to be a market success. I think that's BS, but I wouldn't call myself partisan on this issue as I recently said something similar about Cook's comments regarding the hiding of the ?Watch in the accessories category.
I had the same reaction, but I think he means the iPads are general purpose device, as opposed to other companies that will focus on some aspect of HW because they are focusing on a specific user. That said, I'm not sure I know of any modern tablets that's general purpose, unless you count an eInk-based device as a tablet that focuses on reading books.
In what way have Nexus devices been a flop if the intent is to introduce a new version of Android along with a base hardware platform for it. I think you yourself have noted they aren't meant to compete with licensees nor to be commercial successes, so how do they "flop"?
If a drunk falls down and then says he meant to fall down because he was too far from the ground to start with, is it not a fall?
But two things: first off, Android Enthusiasts predicted strong Nexus sales and continued to think that Nexus devices were not flops after many generation of Nexus flops. I've interviewed lots of Nexus owners, and virtually all of them think that the products sold well.
But now that it is getting impossible to say that Nexus launches have ever been anything other that wildly underachieving flops, Android Enthusiasts are backtracking to say that nothing Google does is ever a flop because the company is just spending billions of dollars trying to show third rate, incompetent hardware makers how they should be making great Android products.
It's like of like communists who say that communism isn't a flop, it's just a wonderful idea that has never been implemented correctly. How many years of failed attempts does it take to prove something is a failure?
Also, if you fact check yourself against reality, you'll find that ever since the first Google Nexus One, the company positioned Nexus as being "a new way to buy a phone," not a way to fail commercially or to not compete with licensees. Also, was Motorola also not "meant to compete with licensees nor to be commercial success"? Because, come on, that's just too much to swallow.
Now turn around and consider how vocal the same Android Enthusiasts were about Apple Maps and iPhone 5c being "flops," even though both have been wildly successful. Just ask Google, which has lost most of its iOS Maps user base and then lost a couple billion dollars trying to compete with Apple's low end phone with its own, unprofitable Moto X flop.
Really, your interlocking circles of logical fallacies are such incredible bullshit that it's hard to even comprehend how you can spew this stuff without hiding behind a phony identity. Oh wait never mind.
I already think it provides a better user experience thanks to the structure of the OS, but no 4.x build is as polished as iOS. I think 5.0 will match the polish of iOS and widen the gap in the sensibility and usefulness that Android already holds over iOS.
You seem to be, but I'm not able to read your mind. Maybe you're just nervous about the software that they bring in tow. Lollipop appears to be poised to finally put Android on equal footing with iOS as far as polish goes. Considering the fact that a large number of iOS users attribute their choice to use iPhones to the polish and predictable performance of the software, I would imagine you might be nervous that iOS's one untouchable distinction might vanish. Thus, without any leg to stand on when criticizing the OS (the lates version anyway), you turn to the devices themselves and compare them in performance benchmarks, which you've always criticized as being unrelated to real usage.
Really? What Continuity and Handoff features has Google built for Mac OS X, Windows and Linux desktop so that you can seamless switch between devices?
HA! Seriously though, you'll have show me what you mean by this Android sensibility that exceeds iOS.
Pot... Kettle. LOL
I had the same reaction, but I think he means the iPads are general purpose device, as opposed to other companies that will focus on some aspect of HW because they are focusing on a specific user. That said, I'm not sure I know of any modern tablets that's general purpose, unless you count an eInk-based device as a tablet that focuses on reading books.
The context: "The problem for Nvidia is that the market for gaming-only tablets is tiny, and its vaunted Tegra K1 with what was supposed to be industry leading graphics is actually slower than Apple's general purpose A8X" indicates that what I mean was that Tegra K1 was supposed to be industry leading graphics, while Apple's A8X is not a dedicated gaming device or an extraordinarily expensive, special purpose niche device. It's going to sell in the tens of millions of units, running everything from enterprise apps to specialized imaging apps to videos games to Facebook.
I already think it provides a better user experience thanks to the structure of the OS, but no 4.x build is as polished as iOS. I think 5.0 will match the polish of iOS and widen the gap in the sensibility and usefulness that Android already holds over iOS.
sensibility "the ability to appreciate and respond to complex emotional or aesthetic influences; sensitivity. (sensibilities) a person's delicate sensitivity that makes them readily offended or shocked"
Yeah maybe that was the right word for "Lollypop" and its new web page appearance.