'that you accuse me of not having an original thought when you clearly know better is sad and unfortunately reflects the fact that you simply dislike me. '
Sloppy again. This is what I wrote "but still how much original analysis is actually going on in his mind ?" and is not what you said I said. It is a fact though that many of your articles have basically been spinoffs of ihub/yahoo debates, nothing wrong with that though but hardly original and groundbreaking ;-).
I don't dislike *you* which is your arrogant way of saying my work is so perfect so that if you criticize it it must be personal. Your work lately has just been terrible, emotionally driven with sloppy facts, analysis and conclusions. Absolutely terrible.
'the original thread can still be read in messages mode.'
apart from Wallis's original post which was a very accurate description of AE's SA/MF MO. No doubt he will repost it.
I also in that thread said use Moorefield as the premium android part because of its gpu. It is clear I copied nothing from your article as you falsely claimed.
'it's interesting that the "big core" cracks x86 instructions into micro-ops while SLM just runs x86 straight through the pipeline.'
Red herring, just an implementation feature. There is a single IBM CISC mainframe instruction that breaks down into thousands of micro-ops, try coding that in RISC lol !
Obviously someone did not like some home truths and went crying to Yahoo and got the original thread deleted ;-). Folks the original thread can still be read in messages mode.
That Ihub thread was interesting as basically it showed I had to convince you that Merrifield/Moorefield were better suited for Android than Bay-T and now you claim I got the idea off you lol ! Seriously it's been obvious to me for sometime that you do not remember importantly what you say and what is said back to you.
aeassa Member Level Tuesday, 02/11/14 10:36:39 AM
Re: mas post# 129568
Post # of 131259
mas, then here's my problem with that. Why is Intel claiming that it's going very broadly with BYT-T to the point of needing contra-revenue for the value segment when Merrifield would make more sense here?
mas Tuesday, February 11, 2014 11:54:05 AM
Re: aeassa post# 129579 Post # of 131259
I'm not sure why you think that. Intel showed off BYT-T running a 64-bit Android kernel and seemed to be really proud of that.
For market differentiation. Moorefield will undoubtedly have a superior gpu to Bay Trail so make that the premium 64-bit part and get the Android profit on that.
'The ARM guys can improve their designs, too, as can Intel - I am arguing that the instruction set architecture isn't a big deal either way. It's about making the smart micro-architectural decisions, the quality of the physical design teams, and even the quality of the compilers and how easy Intel makes it for developers to optimize for their chips. '
When x86 goes 6-8 issue for Core and 3-4 issue for Atom it will be game over for all the Riscs because there is only a limited amount of parallelism you can derive from a processor and ultimately it will come down to how powerful your instructions are with that limited parallelism.
'Let's also not forget that you have continually ignored the fact that there's far more to an SoC than just the CPU. and GPU.'
Yet all you harped on about Bay Trail when it was released was how its gpu was not powerful enough ;-).
'you actually ended up making a post on iHub claiming the very things that I did in my piece. While it doesn't take a genius to actually claim that Moorefield is a better fit for a potential N8 than a Bay Trail-T, the fact is that I had worked through and published this idea well before you made your post.'
Selective amnesia. I have been saying this for months about Merrifield/Moorefield in replies to you and others. So who had the idea first again ?
"mas Tuesday, 02/11/14 05:49:44 AM
Re: aeassa post# 129540
Post # of 131259
'I liked Merrifield until I learned a lot more about when it's actually supposed to be in the market, what modem it was supposed to be paired with, and got some clues as to what kind of design win traction it garnered'
So relaunch it as a tablet chip then. Bet it will be cheaper to make (smaller die) and have fewer BOM issues than Bay Trail."
'I'm also going to take the time to disagree with you now on your claim that because ARM has played its "3-issue" and "6-issue" cards that X86 has more headroom than ARM.'
Please do because it will show who's really the better chip analyst out of us when the future proves me right.
' X86 may indeed have more headroom but this is because as the processors get more complex, Intel can draw upon the "greatest hits" from its big cores to shape the development of its small cores. Pick the most power efficient/sexy features and build a low power core out of them. But it has zilch to do with issue-width.'
WRONG and basically just uneducated waffle.
'Silvermont and Saltwell are both 2-issue, but they are worlds apart in perf/clock and perf/watt even normalized for process. '
Saltwell is in-order and designed for multi-threading, Silvermont is out-of-order, having double the cache and designed in light of Saltwell experience. I have quantified the benefits of out-of-order before as 30-40% so there is no magic to Silvermont's perf.
'you continued to go on about how a discrete modem + apps processor would work just fine.'
Something I still adamantly believe. 28nm SoFIA is a complete waste of time and resources. Who really wants a bigger slower Merrifield ? Seriously, who ?
'Turns out Intel was using "stock browsers" for each device '
LOL, just like the users will be ;-)
'Of course, you then tried to claim that Medfield/CLT+ would be cheap ARM killers when I told you that these SoCs were missing a lot of key functionality/integration that would prohibit their use as long-term weapons against ARM. '
Yeah that's why they are undercutting the cheapest ARM tablets right about now lol. Integration does not mean a damn thing if you can't match the price of discrete parts.
'Further, you've claimed that GPU performance doesn't matter despite a major part of tablet usage is indeed in gaming.'
It's just benchmarketing just like number of cores.
'despite a major part of tablet usage is indeed in gaming. '
'Finally you claim that none of the major handset vendors want to work with Intel as they want to keep Intel out. This directly contradicts Intel's own statements at the Investor Meeting that it would be focusing on the half-dozen smartphone vendors and no longer going after the small players. Do you think Intel has no shot if it has a world-class product? '
Yet they *only* announcedfuture mobile design wins with PC OEMs branching into mobility. I rest my case ;-).
Don't forget the rehashing of the same material over again just with a slightly different angle each time or one new bit of news to create a 'new' article. How many times have we seen the Snapdragon and Merrifield/Moorefield floorplans now ? Also many times I have said something and later on it appears in his article. I don't mind really but still how much original analysis is actually going on in his mind ? Yahoo can put a stop to this hit trolling by showing the author's name in the headline as well so one can avoid the initial click that pays them. That way authors who get a bad rep will pay for it in the end by being ignored.
CURRENT PRICE 3.56, just above resistence, 3.4 ± 0.15, type Single, strength 5
RESISTANCE ABOVE None.
-4.5% at 3.4 ± 0.15, type Single, strength 5
-14.3% at 3.05 ± 0.14, type Single, strength 3
-20.2% at 2.84 ± 0.13, type Single, strength 5
-29.5% at 2.51 ± 0.11, type Single, strength 5
-56.2% at 1.56 ± 0.07, type Double, strength 10
-75.8% at 0.86 ± 0.04, type Single, strength 1
-79.2% at 0.74 ± 0.03, type Triple, strength 3
-82.3% at 0.63 ± 0.03, type Single, strength 1
-83.7% at 0.58 ± 0.03, type Triple+, strength 5
-86% at 0.5 ± 0.02, type Single, strength 1
Confirmation - Strong bullish 3 day chart pattern.
Confirmation - CONFIRMED breakout above 3.55, no resistance in area just above.
Probability - Intermediate trend bearish, Unsustainable uptrend, may turn sideways.
Probability - 1 Day Price change extreme up, may pause/reverse
Probability - Overbought, odds favor short trades.
Probability - Short term typical rally, rally may start to slow.
Confirmation - Extreme 3 day accumulation, but when the stock is overbought this is considered bearish.
Confirmation - Extreme Neutral 1 day moneyflow, on an strong-extreme rally is considered bearish.
' For all you know, X86 ends up being pushed further and further up as ARM continues to march into the mainstream. '
No. ARM has played it three-issue and six-issue cards, Intel has yet to play them for Atom and Core on a superior process, e.g. 10-14nm. Intel Architecture has more potential than ARM going forward, 28nm will be seen as ARM's peak for so many reasons but primarily for design improvement deceleration and transistor cost flat-lining as well as target market saturation.
' there would be no need for SoFIA on TSMC 28nm.'
I don't believe there is a real need for such a product, it's just BK overreacting and doing 'something different' quickly. On 14nm yes I could see how an Airmont pair + integrated modem would make sense when the main product will be quad-core Broxton. On 28nm I don't see what differentiation there is to just adding a Merrifield and 7260 together unless they really cut down the gpu to make the die much smaller.
Those $7 A7s will have bigger things to worry about like how do they compete against 73 sq mm Merrifield when it is eventually used as a performance/price weapon against them.
It's not larger than an A7 or not appreciably so as to counter Intel's density, cost and own foundry advantage. Remember the foundry's transistor cost does not go down below 28nm while Intel's does all the way to 10nm. Intel are going to sell these IoT chips for a dollar, SS stated so in an interview, try getting an A7 for that cost.
'How will Intel compete with ARM at the low end? As an example the Cortex M0 has a gate count of 12,000'
Intel has a process density advantage and it is its own foundry which the Core chips pay for. Atom and Quark can both be sold at or just above die cost if the primary aim is to deprive ARM and its licensees of revenue. This is going to be a generational war as was the original x86-Risc war of the 80s/90s/00s, the one that left Risc to scurry up and down the food chain to enterprise servers and micro-controllers to escape x86.
Current TA says BUY !!
Today the rising medium term moving averages have started the process of collecting all the short term moving averages above them to reform the MA support pyramid. Might take a few weeks for the process to be completed but once it is the stock then advances up again steadily. heslep55/jackiet83 your SHORT is TOAST !!
So Silvermont is at a fundamental disadvantage to A7 in the test that's supposed to fairly test across all platforms. Yep the intellectual lazy sloppiness continues. I tell you what, if you worked for me and produced 'work' like this I would get you fired asap.
'Pay attention to theblueredmonk's post - GB2 was crippled on ARMH
You and he should prove it rather than looking for excuses to ignore it. You think A7's 32-bit score is based on poor optimization ?
' for highlighting the performance results of the only decent low-level CPU benchmark we have'
while totally ignoring how radically different the results are to the previous version of it which should have raised a red flag to you if you was doing your DD instead of rushing over yourself to write yet another sensationalist inaccurate misleading tabloid article. We also have SpecINT2000/6 which is far more 'decent' than geekbench will ever be but of course you keep ignoring that as it would require you spend some real money to do the job properly you pretend to be doing.
'Do note that marsavian himself would often link Geekbench scores'
Indeed GB2 with the proviso that the FP results were buggy. Now the integer results in GB3 are fundamentally flawed.
'It's only when Silvermont came up short that suddenly the results were worthless. '
No, I have criticized GB3 from day 1 after basically seeing it is just compressing,decompressing and encryption in the integer group. Hardly a thorough re[representative suite of integer operations lol.
They look they are being used in the SHA1 test.
The basic point remain, the AES/SHA subtests inflate the A7 integer score in GB3 by over 30% and that is falsely implying that will be the general increase in integer speed across the board because these instructions are enabled. It's just too absurd for words !