Our resident MSc doesn't seem to comprehend what a low level api is.
Ok, I see. Sounds solid. So it'll require extra work, the question is just how much.
And I suppose none of us know how much extra work it'll actually be. Remember there's also been extra work when game devs took use of all the other new features that have been added all the way through ...->DX9->DX10->DX11->... . Games today are a lot more expensive to develop than 10-15 years ago due to that, but the game companies have coped with it.
Or maybe we just put too much trust in linkedin profiles.
I tried to verify the profile and never found anything.
True, we will have to see if turns out to be true. Still, it wouldn't be that surprising to see Intel HR give that guy a stern talking to (if not fired the guy) for posting confidential info and so he took down his linkedin.
Or maybe we just put too much trust in linkedin profiles.
I tried to verify the profile and never found anything.
If Intel doesn't counter the Zen "threat" with an 8 core "mainstream" chip, then I will be truly and genuinely surprised.
That's kind of my feeling (hope) as well. Here's hoping that Intel finally releases mainstream CPUs with 6-8 cores in them!
Might actually get me to buy a high-end mainstream CPU again.
Also, slightly OT, but I was thinking, what with Anandtech's review of 8-core mobile phones, and coming to a conclusion that browsers like Chrome could utilize that many threads, and do it at a lower power consumption than the equivalent-performance higher-clocked quad-cores, then I had to wonder, will Intel release 8-core consumer Atom derivatives any time soon?
I mean, Z3735F and friends were already quad-core, which was fairly impressive for a, what, 6-10W TDP chip, but I'm wondering, if at 10nm, we might see octocore mobile Atom CPUs?
ARM promoted big.LITTLE, not octa cores. It's ARM customers such as Mediatek and Samsung that started that silliness.Octa cores are a way for ARM Holdings to collect more royalties per chip. Kudos to ARM, they know what they're doing, but from a technical perspective octa cores are useless in a phone.
My logic comes down to this: DX12 requires a non-trivial increase in the amount of work/effort compared to DX11 because it gives developers a lot more control.
This shouldn't be a problem for big, rich studios, so Call of Duty, Battlefield, etc. all of the big names will have no issue transitioning to the new API. But for small indie developers or even mid-sized ones on a relatively tight budget, I can imagine that they may forgo the benefits associated with DX12 for now in order to be able to get their games out to market in a reasonable time-frame and at reasonable development costs.
Longer term, devs may have no choice but to migrate to DX12 though.
ARM promoted big.LITTLE, not octa cores. It's ARM customers such as Mediatek and Samsung that started that silliness.
What we see in the use-case analysis is that the amount of use-cases where an application is visibly limited due to single-threaded performance seems be very limited. In fact, a large amount of the analyzed scenarios our test-device with Cortex A57 cores would rarely need to ramp up to their full frequency beyond short bursts (Thermal throttling was not a factor in any of the tests). On the other hand, scenarios were we'd find 3-4 high load threads seem not to be that particularly hard to find, and actually appear to be an a pretty common occurence. For mobile, the choice seems to be obvious due to the power curve implications. In scenarios where we're not talking about having loads so small that it becomes not worthwhile to spend the energy to bring a secondary core out of its idle state, one could generalize that if one is able to spread the load over multiple CPUs, it will always preferable and more efficient to do so.
In the end what we should take away from this analysis is that Android devices can make much better use of multi-threading than initially expected. There's very solid evidence that not only are 4.4 big.LITTLE designs validated, but we also find practical benefits of using 8-core "little" designs over similar single-cluster 4-core SoCs.
According to AnandTech it's not silliness but the best approach:
What approach a specific company chooses does not prove anything. The technical evidence and principles in the article is valid regardless. It's generic and company-agnostic...I am sure AT knows more than Apple/Qualcomm do
Octa cores are stupid marketing gimmicks.
... in the same way that quad-cores, are "stupid marketing gimmicks" too, right?
All we need are uber-clocked single cores. Well, except for that pesky thing called power consumption and power efficiency... it's not by accident that the Z3735F is a 1.33Ghz / 1.86Ghz quad-core, rather than a 7Ghz single-core.
Intuitively it might seem so. But reality shows a completely different picture. You really should read the AT article I referenced, at least the quote from it that I just posted, because it totally contradicts that quad or even octa cores should be considered a marketing gimmick in mobile phones. I know it may sound strange, but the technical evidence is all there.In phones, quad cores are stupid marketing gimmicks.
What approach a specific company chooses does not prove anything. The technical evidence and principles in the article is valid regardless. It's generic and company-agnostic...
Also, Qualcomm has used b.L and 8 cores too. So does Samsung and Mediatek. That constitutes the majority of the smartphone market share.
Haha
Oh God! The article doesn't prove a damn thing. Seriously. It its useless to the discussion/point you are trying to make. All the article demonstrates is that Android can schedule a lot of threads. It provides no evidence nor makes no conclusion that ramping up lots of cores is better in either performance or power.
And less than 10% of the profit of the smartphone market...
And FYI, Qualcomm has never willingly used b.L, they have only done it out of nessecity for 1 generation because their management fubar'd and now that they are past that fubar point, they aren't using it going forward.
Ok, so you've figured it all out and the Anandtech reviewers are stupid and got everything wrong. Right. We get it.
It's you who are saying that Anadtech reviewers are stupid and got everything wrong because one company (Apple) has chosen another path than the article concludes is the most optimal one from a technological perspective.Its you trying to convince us all that Qualcomm and Apple is run by idiots because you found an article on AT you dont fully understand.