Don't believe so? Only mention I can find him making of Haswell overclocking is this post on twitter. I guess you can take the statement of "Haswell = Overclocking" with a link to a Tom's Hardware article discussing the just-revealed details on Haswell overclocking features to imply that it's an awesome overclocker if you want to... I tend to just take it as a statement that Haswell includes more overclocking features.
From Francois on Facebook, in reaction to article in link below:
So, let's call a cat, "A CAT" ... there are now people "lying" and providing financial advised, if I were the guys about to lose money on some of those stocks, I would take this firm to court : "we have seen similar benchmarks which clearly score the 1.8GHz Cortex A12 (ie not even the high end A57) ahead of the 2.4GHz Silvermont"
This is a pure lie, the silicon of A12 does not exist, and when it comes to reality, it will not win either. Gareth Jenkins is purely lying right there. This is my personal opinion, my employee did not ask me to post anything here, this is just outrageous, and nobody call them on their ignorance or dishonesty.
http://blogs.barrons.com/techtrader...icensing-overblown-says-ubs/?mod=yahoobarrons
If cherry-picked ICC-biased benchmarks are the best Intel has to show, is it any surprise that ARM doesn't seem worried?
OMG,ARMIntel is doomed. /s
Can we save the 100 page thread for when some real meaty information shows up?
And how can Francois (Piednoel?) know that Cortex-A12 "will not win either" if no A12 silicon exists. He surely didn't even see the full specification of Cortex-A12 as the information is not public. I wonder who is lying there.From Francois on Facebook, in reaction to article in link below:
http://blogs.barrons.com/techtrader...icensing-overblown-says-ubs/?mod=yahoobarronsSo, let's call a cat, "A CAT" ... there are now people "lying" and providing financial advised, if I were the guys about to lose money on some of those stocks, I would take this firm to court : "we have seen similar benchmarks which clearly score the 1.8GHz Cortex A12 (ie not even the high end A57) ahead of the 2.4GHz Silvermont"
This is a pure lie, the silicon of A12 does not exist, and when it comes to reality, it will not win either. Gareth Jenkins is purely lying right there. This is my personal opinion, my employee did not ask me to post anything here, this is just outrageous, and nobody call them on their ignorance or dishonesty.
Btw, on Silvermont's graphics -
Intel claims a 3x improvement over Clovertrail, which to me is likely and realistic. However you have to understand the incredibly low performance they are coming from to see what that means.
http://www.anandtech.com/show/7082/...-preview-qualcomm-mobile-development-tablet/6
Even with 3-4x improvement, Intel will still be only half as fast as the S800's Adreno 330, and still lagging far behind Jaguar. So even if the cpu is faster, it'll still feels more sluggish than the main competitors in graphics.
It's still an Atom, remember that.
Francois didn't make the "overclocker's dream" comment, that was supposedly and intel engineer on reddit.
He did, however, claim the toms hardware preview of Haswell wasn't intel sanctioned and that the numbers were inaccurate (everyone was upset about how it didn't show much improvement over ivy). Tom's later review pretty much showed exactly the same results.
-maybe intel has links to PRISMAnd how can Francois (Piednoel?) know that Cortex-A12 "will not win either" if no A12 silicon exists. He surely didn't even see the full specification of Cortex-A12 as the information is not public. I wonder who is lying there.
Toms Preview was inaccurate. It had bandwidth issues.
What precisely makes you think that I don't? But at least the Intel performance projections tell us what they're based on. The Silvermont vs Saltwell chart compared the new atom against a Z2580 with a particular suite of benchmarks. The Silvermont vs ARM charts compare only on SPECint rate_base200 against "expected configurations of main ARM*-based mobile competitors using descriptions of the architectures" - aka the only thing I found interesting there was the 1W for DC and 1.5W for QC power consumption that they used for the comparisons. (Since we don't know what architectures they're comparing against and the benchmark isn't a good metric for much.) Regardless, it's still far more information than I've seen ARM give to justify their claims.I like your thinking, however you need to apply it to Intel and not ARM.
Intel already knows how the S800 performs and they chose to show biased and cherry-picked benchmarks as a response. That's what the loser usually does.
Definitely. Good thing that 3D graphics is used for nothing other than gaming eh? And the most popular games for smartphones/tablets have extremely low requirements in that regard. Now sure for the value notebook segment it leaves an opening for Jaguar, but that's the compromise Intel chose to make in order to make Baytrail ideally suited for its intended market.Even with 3-4x improvement, Intel will still be only half as fast as the S800's Adreno 330, and still lagging far behind Jaguar. So even if the cpu is faster, it'll still feels more sluggish than the main competitors in graphics.
Because even ARM doesn't claim that A12 will be faster than A15 and he almost certainly already knows how fast A15 products are? That's not to say A12 isn't impressive, but the only areas where it might beat Silvermont are density and then power efficiency at certain points of operation.And how can Francois (Piednoel?) know that Cortex-A12 "will not win either" if no A12 silicon exists. He surely didn't even see the full specification of Cortex-A12 as the information is not public. I wonder who is lying there.
What precisely makes you think that I don't? But at least the Intel performance projections tell us what they're based on. The Silvermont vs Saltwell chart compared the new atom against a Z2580 with a particular suite of benchmarks. The Silvermont vs ARM charts compare only on SPECint rate_base200 against "expected configurations of main ARM*-based mobile competitors using descriptions of the architectures" - aka the only thing I found interesting there was the 1W for DC and 1.5W for QC power consumption that they used for the comparisons. (Since we don't know what architectures they're comparing against and the benchmark isn't a good metric for much.) Regardless, it's still far more information than I've seen ARM give to justify their claims.
Definitely. Good thing that 3D graphics is used for nothing other than gaming eh? And the most popular games for smartphones/tablets have extremely low requirements in that regard. Now sure for the value notebook segment it leaves an opening for Jaguar, but that's the compromise Intel chose to make in order to make Baytrail ideally suited for its intended market.
Because even ARM doesn't claim that A12 will be faster than A15 and he almost certainly already knows how fast A15 products are? That's not to say A12 isn't impressive, but the only areas where it might beat Silvermont are density and then power efficiency at certain points of operation.
If it is true, then it was properly running at stock 1.1Ghz and running in Turbo mode during benchmarks which means a dual core 2+Ghz would be faster then Snapdragon 800.
Definitely. Good thing that 3D graphics is used for nothing other than gaming eh? And the most popular games for smartphones/tablets have extremely low requirements in that regard. Now sure for the value notebook segment it leaves an opening for Jaguar, but that's the compromise Intel chose to make in order to make Baytrail ideally suited for its intended market.
The above makes it sound like you're crediting the 'best tablet experience' of the ipad to its graphics? If so, that's quite incorrect. The difference in 'experience' is due entirely to the software platform, not the hardware. Basically the facts that the best selling tablet is the ipad and the ipad SoC has massive graphics do not combine into a conclusion of the ipad is the best selling tablet because its SoC has massive graphics. (Logical fallacy.)I dont find this as good excuse for below average gpu perf (although I don't know if that is true/false of this chip, just in general) apples socs are proof that the gpu is huge for mobiles, just look at the a6x, that chip is huge, 123mm^2 [tegra 4 ~80mm^2] and even though the tegra 3 had more cpu perf, no one really cared, because the experience on the ipad is so much smoother, the pvr sgx mp4 is even overpowered and it gives the best tablet experience.
Since Qualcomm uses a custom design, what IPC improvements precisely are 'available from ARM right now'? As well, the move to 20nm won't help them much at all, it's primarily a density improvement with very little change to the electrical characteristics of the transistors. The subject of modems is actually one of the primary points of interest for the adoption of Silvermont into smartphones - Intel should actually be able to provide a comparable in-house solution to Qualcomm for the first time.I think by doing a few more IPC improvements ( which are available from ARM right now ) and a move to 20nm will make them competitive if not better then Intel on both Power and Performance.
Pricing for Snapdragon will still include everything from LTE and other Wireless connectivity.
So i still dont see how Intel is having a clear advantage here until they move to 14nm, which will be against others 16nm FinET counter parts.
The above makes it sound like you're crediting the 'best tablet experience' of the ipad to its graphics? If so, that's quite incorrect. The difference in 'experience' is due entirely to the software platform, not the hardware. Basically the facts that the best selling tablet is the ipad and the ipad SoC has massive graphics do not combine into a conclusion of the ipad is the best selling tablet because its SoC has massive graphics. (Logical fallacy.)
Since Qualcomm uses a custom design, what IPC improvements precisely are 'available from ARM right now'? As well, the move to 20nm won't help them much at all, it's primarily a density improvement with very little change to the electrical characteristics of the transistors. The subject of modems is actually one of the primary points of interest for the adoption of Silvermont into smartphones - Intel should actually be able to provide a comparable in-house solution to Qualcomm for the first time.
Oh, and even if TSMC/GF deliver on their promised timelines and performance metrics for their 16nm FF/14nm XM processes all indications are that they are merely going to be comparable to Intel's 22nm process. When you actually look at the marketing claims Intel's process lead is only going to increase in the near term... and Intel's the only one who has actually been meeting their marketing claims for the past few years in the first place.
I am very certain that the software needs hardware in order to function, this might be a chicken or the egg situation but I am sure that it is the hardware that allows the software to dictate that experience to the user...
I am very certain that the software needs hardware in order to function, this might be a chicken or the egg situation but I am sure that it is the hardware that allows the software to dictate that experience to the user.
The hardware dictates functionality and feature set, [high graphics quality, long battery life etc], software is just the implementation...In any case I am not arguing for who the best platform is [wrong choice of words I guess].
I guess all I am saying is if the best selling tablet platform, has been using relatively oversized gpus in their socs and not so much cpu[quad? octa?], maybe they know something we dont...
OMG ARM is failing!!!!
How would consumers tolerate their ARM crap right now when this new chip would load their Facebook imperceptibly faster?
The biggest problem with ARM is the fact that it`s not powerful enough for Flash and therefore YouTube.
It probably really was running at 1.1GHz.
In AnTuTu Intel also scores much higher with a 2GHz 2C4T Saltwell than S600 does with 4 1.7GHz Krait 300 cores.
Exactly. I don't get the APU hype on desktop and even less the GPU hype in smartphones. And if you play games on a smartphone or tablet you basically are already admitting that graphics matter jackshit to you...So I agree. It's good enough. I rather pay less money for the phone and get better battery life and CPU performance than a huge GPU. CPU performance actually matters a lot when browsing on anemic smartphone CPUs.
The biggest problem with ARM is the fact that it`s not powerful enough for Flash and therefore YouTube.
So, in AnTuTu quad 2-wide out of order Silvermont cores will perform better than dual 4-wide out of order Haswell cores with hyper threading.
According to the following link, a Core i5-4200U clocked at 1600 MHz (2600 MHz Turbo Frequency) scores 54,861 in AnTuTu:
http://vr-zone.com/articles/samsung-ativ-q-shatters-antutu-record-core-i5-scores-over-54000-points/40677.html
Given the excellent scaling of the AnTuTu benchmark, a quad core Silvermont running at 1600 MHz will score over 63,000 points in AnTuTu rendering it 15% better than a Haswell i5-4200U.
The excellent scaling of the AnTuTu benchmark appears to be less than credible, but a quad core Silvermont running at 2100 MHz will score over 82,000 points in AnTuTu. This would render it 50% better than a Haswell i5-4200U (Tray price $287).
Interesting, but doesn't this look strange?