:thumbsup:Because slow-ass crappy computers are annoying.
A cell phone is not just a cell phone, it is a portable web browser, game machine, and general use computer. Would you compromise performance on your desktop?
but do you think you could tell the difference between them[special benchmark mode/regular mode] if you were to use them as let's say...cellphones?Because it's stupid to only boost speed for benchmarks. It's misleading at best and outright lying if you're cynical.
So besides cheating in benchmarks, they are also lying about the S4 ACTIVE being waterproof....
Read through it again and from what I gather, samsung default underclocks their cpu and gpu but on certain benchmarks they lift the clocks...the chip is capable, they are not going above/beyond their spec, so why is this so bad? they are handicapped in most scenarios save for certain benchmarks where you would want to see performance.
Thinking about this some more, I guess they're trying to hide an architectural problem with big.little. I can only assume that before you run the benchmark, everything is sitting around on the A7. Once you start the benchmark, there's probably a 1 time performance hit copying over the states from the A7 cores to the A15 cores and therefore hurting the score. So they're "warming up" the processors prior to the benchmark to hide any latency problems they have with core swapping.
Let's pick some very pessimistic numbers and say cache flushing requires 200 cycles per 32-bit word (for 1 GHz, that'd mean 20MB/s which you'll agree is *way* below what current SoC/RAM do). I'll let you compute how many cycles it takes to flush 4x32KB L1 Dcaches and the 2 MB L2 cache, I'll even let you do this as if all of the caches contained dirty lines. Yes, that's it, ~100M cycles for a CPU that runs at more than 1 GHz.The first "octa-core" actually suffers from this especially- something about the switching mechanism is utterly borked, and has been disabled. As such they have to completely flush the caches when they switch from the A7 to A15 clusters.
The first "octa-core" actually suffers from this especially- something about the switching mechanism is utterly borked, and has been disabled. As such they have to completely flush the caches when they switch from the A7 to A15 clusters.
I remember a post on the forums talking about how flimsy the whole waterproofing method looked.
http://forums.anandtech.com/showpost.php?p=35110710&postcount=4
I would prefer something that's meant to be water resistant to be a little more idiot proof. In the example of the reddit post above, sure he got the back plate all secure but WOOPS, forgot to tighten the USB cover. /dead device.
they are handicapped in most scenarios save for certain benchmarks where you would want to see performance.
:thumbsup:
Strange, why this needs to be explained. Some spend hours of every day for work/ consumption and upgrade every 6-12 months. Should some be spending as much money/monthly as they do? If not contributing to work. Well that's a different discussion and then it gets down to also being a hobby/ their web enjoyment 'toy'.
Clocks are a fine line and would be set that high if the manufacturer could. But are probably limited by cooling/ battery life.
but do you think you could tell the difference between them[special benchmark mode/regular mode] if you were to use them as let's say...cellphones?
It is funny to see how people dog samsung for making their chips run at non-undercloked speeds, when the performance is so high.
Seriously lol.
Funny how on a tech forum:
User 1: OMG HASWELL ISN'T FAST ENOUGH!?
User 2: What you need it to be faster for?
User 1: BENCHMARKING!!!!!
But when it comes to phones, people claim to care less about performance. Maybe because some people are simply antisocial and never need to use their phones, but those of us who need to use it, phone performance is important.
Considering I send more texts than my whole family combined, and I'm almost always on it (Sometimes I accidently google things on my phone AT my computer), phone benchmarks are important to me.
Kind of pisses me off actually that Samsung releases 2 different versions (US/International) of their phones, but usually only talks about the specs of the international phone, thus you think you're getting the fast international til you realize you get the watered down US phone.
It does look like a broken "boost" type of functionality. Seriously though, what good is a boost function that only works for benchmarks and the camera app?
Oh well, I guess if you are willing to root you can probably pick that extra speed back up easily enough...
you've got me there, I would be irate if my highend cpu doesnt match up to advertised speeds[like isps?].Seriously lol.
Funny how on a tech forum:
User 1: OMG HASWELL ISN'T FAST ENOUGH!?
User 2: What you need it to be faster for?
User 1: BENCHMARKING!!!!!
But when it comes to phones, people claim to care less about performance. Maybe because some people are simply antisocial and never need to use their phones, but those of us who need to use it, phone performance is important.
Considering I send more texts than my whole family combined, and I'm almost always on it (Sometimes I accidently google things on my phone AT my computer), phone benchmarks are important to me.
Kind of pisses me off actually that Samsung releases 2 different versions (US/International) of their phones, but usually only talks about the specs of the international phone, thus you think you're getting the fast international til you realize you get the watered down US phone.
Again, just amazing to me. If Intel released Haswell and you bought a 4770k, only to find out that it never ran at advertised speed, and only ran at turbo speed during benchmarks, and not even at gaming, you'd be irate.
The only reason I can think of that people are like "Meh whatever" is that they rarely use their cell phones.