Ultra Thin Notebooks Climbing While Desktop Sales Decline; Tablets Slowing

[InvalidError]

It has a 5.7in screen with lower res at 1080p, vs. shield at 8in 1920x1200. Last time I checked 5.3hrs is 35% longer life than 3.9hrs not 10% you claimed,

Did you actually read the graphs or did you blindly pick up the best-looking bar with Nvidia next to it and assume it was the K1?

The NVIDIA Shield (notice the missing "Tab" - that would be last year's Tegra 4 handheld) scored 5.3h but the Shield Tab scored 4.3h with GFXBench locked at 30fps. 4.3h is ~10% more than 3.9h.

If the K1 had not been capped to 30fps because "most games shouldn't come close to stressing the Kepler GPU," the K1's battery score would have been even worse: if you look dead last on the graph, without the 30fps cap, the Tegra Tab sinks to only 2.2h.

As for the resolution difference, 1920x1080 vs 1920x1200... not a big deal.

Bottom line: using K1's full power nukes batteries like nobody's business... the SoC uses close to 10W under full load.

 

[somebodyspecial]

It has a 5.7in screen with lower res at 1080p, vs. shield at 8in 1920x1200. Last time I checked 5.3hrs is 35% longer life than 3.9hrs not 10% you claimed,

Did you actually read the graphs or did you blindly pick up the best-looking bar with Nvidia next to it and assume it was the K1?

The NVIDIA Shield (notice the missing "Tab" - that would be last year's Tegra 4 handheld) scored 5.3h but the Shield Tab scored 4.3h with GFXBench locked at 30fps. 4.3h is ~10% more than 3.9h.

If the K1 had not been capped to 30fps because "most games shouldn't come close to stressing the Kepler GPU," the K1's battery score would have been even worse: if you look dead last on the graph, without the 30fps cap, the Tegra Tab sinks to only 2.2h.

As for the resolution difference, 1920x1080 vs 1920x1200... not a big deal.

Bottom line: using K1's full power nukes batteries like nobody's business... the SoC uses close to 10W under full load.

LOL... Grabbed the wrong number, but it's still the same story just a bit better for your case (not much though). 2.2 inches less screen and lower res on top. Nothing uses K1 full power, that is the point of the article and why Joshua re-visited it. What runs under full load besides a benchmark designed to do it? Nothing. Hence the revisit to prove real world usage. Capped at 30fps it was still better than everyone else, so I don't get your point. You're ignoring his quote:
"By capping T-Rex to 30 FPS, the SHIELD Tablet actually comes quite close to the battery life delivered by SHIELD Portable with significantly more performance. The SHIELD Portable also needed a larger 28.8 WHr battery and a smaller, lower power 5" display in order to achieve its extra runtime. It's clear that the new Kepler GPU architecture, improved CPU, and 28HPm process are enabling much better experiences compared to what we see on SHIELD Portable with Tegra 4"

Also ignoring this, when capped at 30fps which is still BETTER than s800:
"As one can see, despite performance near that of the Adreno 330, the GPU in Tegra K1 sits at around 450 MHz for the majority of this test."

Note that is 400mhz off it's 852 rated speed (did you READ the article text?). So it can match the NOTE 3's gpu (beat it actually), and do it with nearly 1/2 clocks. How do you think that battery life would be in this case? VERY good. How do you think battery on note 3 would do with another 2.2inches on the screen and a slightly higher 1920x1200 res on top? WORSE than they are showing. You're comparing a phone (large one, but a phone) to a tablet and calling it even?

"After all, Tegra K1 delivers immense amounts of performance when necessary, but manages to sustain low temperatures and long battery life when it it isn't."

It's like you have blinders for everything he says that is important. LONG battery life doesn't sound like a jab in NV's face. It sounds good. I picked the wrong bar before (oops), but that hardly changes the story or the message of the article. I don't know why you attempt to take it back into NON-REALISTIC territory again, which is opposite of the point of the article. You're acting like K1 runs full out like a benchmark designed to do that all day which he CLEARLY is stating doesn't happen. And when you do that, how badly does it beat the compeitition? ROFL. K1 rocks, no matter how much you dislike NV. And note also he mentions maxwell will be even better at all these things at the end of the article. Again, I say good luck to qcom.

AMD needs to get in this game too (like yesterday soon), while both have the gpu advantage/drivers/game devs experienced with their hardware for 20yrs etc. There is room for both in the tablet war but not if AMD waits another few years to get in it. There are billions at stake for both of these, and can easily take some qcom/apple/samsung soc money with leading gpus as gaming comes into it's own on AndroidL, 64bit etc. HTC Nexus 9 will be the start of the changing tide if not before it already (google's stuff gets noticed) 😉

 

[somebodyspecial]

oops, meant 2.3 where I said 2.2in. NV's is a full 8in, so your phone you're comparing is 2.3in smaller, I kept hitting a 2, so an even larger diff than I said (but you get the point anyway).

 

[InvalidError]

Nothing uses K1 full power, that is the point of the article and why Joshua re-visited it. What runs under full load besides a benchmark designed to do it? Nothing.

Answer this: what is the point of having a K1 if you cannot use more than 30-40% of its processing power without nuking the battery?

Between the Shield Tab capped at 30fps lasting 4.3h and the one running full-bore lasting 2.2h, the only thing that changes is the SoC's power draw and that increases by about 4.3W.

While most current games may not use anywhere near that much, games designed for the K1 or with the K1 in mind will likely use a fair more than the current average. After all, what is the point of having a super-powerful SoC if nothing actually uses it?

 

[somebodyspecial]

Nothing uses K1 full power, that is the point of the article and why Joshua re-visited it. What runs under full load besides a benchmark designed to do it? Nothing.

Answer this: what is the point of having a K1 if you cannot use more than 30-40% of its processing power without nuking the battery?

Between the Shield Tab capped at 30fps lasting 4.3h and the one running full-bore lasting 2.2h, the only thing that changes is the SoC's power draw and that increases by about 4.3W.

While most current games may not use anywhere near that much, games designed for the K1 or with the K1 in mind will likely use a fair more than the current average. After all, what is the point of having a super-powerful SoC if nothing actually uses it?

The point is it's there when the games come. Also as reviewers have noted, and I myself would do, you'll probably plug in most of the time and use it out to tv (where there's always a handy plug available). I'll be playing android games from my couch, not on any tablet or shield unit. Only OUT of the house is battery a concern to me, and I won't be out that long (sitting in doctors office etc).

Again you're quoting MAXED out gpu numbers though, no game does that the entire time. So those hour estimates go up. I highly doubt even next xmas (2015) will tax this thing to death as that will not be the lowest common denominator for many. It may be more than a year before they really start aiming DIRECTLY at K1, except for special NV ports etc to show off the hardware power. Current games are barely beginning to aim at T4 like power and ONLY a very few (modern Combat 5?, not many others). This years xmas games and most of the next years games will be aimed at T4/S800 levels, not more again except in rare cases.

The main point is, the other guys have NO OPTION here. They don't have the power. I'd rather have the option to plug in when desired LATER when games come that use that power and play all day on my tv. If you buy ANY other soc, you're just out for those LATER games. Plugging in won't give them the perf to get that job done at all right?

You'd rather NOT have that power ability when desired? OK. Buy a Qcom 😉 Basically the same as a K1 in battery but nothing later when killer games come, you'll be left out or buying a new one

 

[somebodyspecial]

We're also talking THIS gaming device. Clearly K1 can get 11.5-13hrs in a chromebook, which is better than Intel's 8.5hr in the same companies Acer chromebooks. Baytrail has worse battery than K1 in comparable chromebooks, ALL from acer. Baytrail has larger battery (3950mah), smaller screen 11.3in, lower res (1366x768) and yet loses to K1 with 3220mah, 13.3in screen and 1080p with 4GB of mem vs. Baytrail's 2GB. You're blowing battery life out of proportion here.

You can feel free to check the datasheets for the chromebooks at acer's site.

Copied from my anandtech post - Not directly related but makes the POWER point clearly:
http://us.acer.com/ac/en/US/content/model-datashee...
Acer with Haswell 2955 says 8.5hr for 3950mah, but only 11.6in and 1366x768.

http://us.acer.com/ac/en/US/content/model-datashee...
Acer K1 13in with 1080p, 3220mah but 11.5hr. I think haswell 2955 loses based on the specs.

I think it's safe to say Intel lost. Smaller screen, lower res AND far bigger battery but 3hrs off the battery life? That's a loss right? So 2955u means nothing.

http://us.acer.com/ac/en/US/content/model-datashee...
$299 2955u model has 7.5hr life and a larger psu (65w vs. K1 45w). Same 2GB 32GB ssd (as the higher K1 model), so not sure how to explain this other than Intels sucks more power.

http://us.acer.com/ac/en/US/content/model-datashee...
4GB/32GB SSD K1 13in model also 11.5hr 45w psu. Intel sucks more power right?

Intel, PowerVR and Qcom all seem behind now. Denver should improve things more in Nov as it drops the 4+1 A15, for plain dual core@2.5ghz and is in house or NV has failed at making a cpu 😉 I'm guessing success or google wouldn't be going with them for HTC Nexus 9 in nov.

 

[natcparis]

Slowing? say it fast though.