Apple A9X Geekbench

[raghu78]

I don't understand why the view is "OMG INTEL SUX" rather than "Wow, Apple did a damn fine job" with its CPU?
Intel is a merchant chip vendor that is fundamentally in a different business than what Apple is in. Apple makes a lot of money selling fully functional devices; SoC cost is an important but relatively small part of the bill of materials of an Apple product.

SoC cost is literally everything for Intel; it needs to make sure that the die sizes of its products are cheap enough to produce in order to support many customers offering products at many different price points.

If you see how the die size of Intel dual and quad cores have been steadily shrinking from the 32nm generation its clear that Intel is able to get away with it because AMD has failed miserably. Meanwhile Intel gross margins have been rising all the time and Intel has been more keen on their gross margins than providing more performance/cores for the same price. So I am happy Apple is challenging Intel atleast in the mobile market.

Apple has achieved something truly excellent here with the A9 and the A9X -- these are wonderful chips and are the product of first class engineering and a laser-like focus on a handful of great products. But to criticize Intel because Apple managed to deliver an SoC with competitive single-threaded performance doesn't seem fair to me.

I want Intel to be challenged strongly in all its markets. Right now its all one way traffic for Intel in servers, desktops, notebooks. In the mobile market Intel is not dominant and faces fierce competition from the ARM ecosystem. In fact I am a strong believer that ARM ecosystem will eventually provide good competition to x86 across the computing spectrum from phones/tablets to notebooks/desktops and servers. Apple is proving it can be done and I expect hopefully there are other companies who do it in other segments of the computing industry.

 

[Arachnotronic]

If you see how the die size of Intel dual and quad cores have been steadily shrinking from the 32nm generation its clear that Intel is able to get away with it because AMD has failed miserably. Meanwhile Intel gross margins have been rising all the time and Intel has been more keen on their gross margins than providing more performance/cores for the same price. So I am happy Apple is challenging Intel atleast in the mobile market.

Dude, Intel is a business that's run for the benefit of its stockholders. Obviously they're going to try to maximize the profits they deliver to their shareholders, just as Apple tries to do the same for theirs! I mean, FFS, they are charging $99 for a stylus. Why didn't Apple just bundle it in with the iPad Pro for the benefit of its customers?

Do you honestly think Apple couldn't have "given" customers more with the iPhone 6s/6s Plus at the expense of its margins?

 

[witeken]

Do you honestly think Apple couldn't have "given" customers more with the iPhone 6s/6s Plus at the expense of its margins?

Yes. The 16GB is just downright pathetic.

 

[stingerman]

If Apple released a high performance notebook version of the A9, it would be on a more mature process that can handle higher speeds. The micro-architecture design is competitive. Me thinks they'll wait to 2017 to replace Intel with their own.

 

[Arachnotronic]

Yes. The 16GB is just downright pathetic.

From a customer perspective yes, but from a business perspective it is downright brilliant and exactly the kind of shareholder-friendly move that Apple should be doing.

 

[stingerman]

Yes. The 16GB is just downright pathetic.

I used to think that until I took all my devices to iCloud Photo, all of a sudden, the 16GB devices always have enough space...

 

[raghu78]

Dude, Intel is a business that's run for the benefit of its stockholders. Obviously they're going to try to maximize the profits they deliver to their shareholders, just as Apple tries to do the same for theirs! I mean, FFS, they are charging $99 for a stylus. Why didn't Apple just bundle it in with the iPad Pro for the benefit of its customers?

Do you honestly think Apple couldn't have "given" customers more with the iPhone 6s/6s Plus at the expense of its margins?

I am looking at it from a consumer's perspective and not a stockholder. I would be glad if Android high end phones and tablets challenged Apple iphones and ipads better and drove their pricing lower or made Apple give more performance / features at the same price. I understand that every company tries to maximize profits but I am more bothered about the state of the computing industry than a single company.

 

[asendra]

Yes. The 16GB is just downright pathetic.

I'm not excussing them, least of all at the prices they charge, but

A) A lot of users don't really need more, so it makes sense for them to have a 16gb tier.
B) It makes you wonder what kind of impact would it have globally for the best sold device in the world, which amounts for a pretty non-negible % of global memory shipments, to double it's memory usage.

 

[Arachnotronic]

I am looking at it from a consumer's perspective and not a stockholder.

Right, I get that, but to understand why things are they way they are you need to take the investor's point of view.

I would be glad if Android high end phones and tablets challenged Apple iphones and ipads better and drove their pricing lower or made Apple give more performance / features at the same price. I understand that every company tries to maximize profits but I am more bothered about the state of the computing industry than a single company.

Android phones pack in lots of features, some of them arguably pack in more features than Apple does. But Apple sells more than just raw specs, they sell a software ecosystem, a world class retail experience, amazing customer support, and a brand that carries a lot of value.

Hell, when I look at my iPhone 6s Plus, the mere fact that there is a shiny, really slick Apple logo on the damn thing makes me "value" the device more than I would without it. Apple's brand is that strong.

There is a lot that goes into building a truly great business and it goes far beyond CPU/GPU benchmarks

 

[Arachnotronic]

I understand that every company tries to maximize profits but I am more bothered about the state of the computing industry than a single company.

The computing industry is in wonderful shape. The kind of performance that you can buy on the cheap today is nothing short of a miracle and it is only getting better.

Even for the $950 that Apple charges for a fully loaded iPhone 6s Plus, can you honestly say that's a bad deal given how useful and well-built the thing is? I can check my emails, play games, play music, watch videos, and so much more. And my experience while doing so thanks to the processing power + well made OS is nothing short of great.

And, next year, I'll be able to sell the thing for at least $500 (probably more) to finance the purchase of an iPhone 7 which will be better in every respect than the 6s.

We live in magical times; don't be so cynical

 

[raghu78]

The computing industry is in wonderful shape. The kind of performance that you can buy on the cheap today is nothing short of a miracle and it is only getting better.

Even for the $950 that Apple charges for a fully loaded iPhone 6s Plus, can you honestly say that's a bad deal given how useful and well-built the thing is? I can check my emails, play games, play music, watch videos, and so much more. And my experience while doing so thanks to the processing power + well made OS is nothing short of great.

And, next year, I'll be able to sell the thing for at least $500 (probably more) to finance the purchase of an iPhone 7 which will be better in every respect than the 6s.

We live in magical times; don't be so cynical

The mobile market is fiercely competitive. The market which cries out for competition is the one which Intel has monopolized - desktops, notebooks and servers. Thats where I want ARM ecosystem to make a big impact. :thumbsup:

 

[Arachnotronic]

The mobile market is fiercely competitive.

Getting less competitive, though. Economics say that you can't have a bazillion chipmakers all gunning for their piece of a pie that is suddenly not growing so fast. In fact there are only two really viable merchant vendors in mobile: Qualcomm and MediaTek. Qualcomm is very clearly ahead of MediaTek in key technologies, to boot.

In fact, the market is getting to the point where heavyweights like Qualcomm have to lay off significant personnel and cut back on investment just to make the economics work. Competition might be good for the consumers, but just ask the 15% of QCT employees that are jobless now, as well as Qualcomm stockholders who are now dealing with a share price in the gutter, just how "great" competition is

The market which cries out for competition is the one which Intel has monopolized - desktops, notebooks and servers. Thats where I want ARM ecosystem to make a big impact. :thumbsup:

The competition in desktops happened long, long ago. If you are a student of this market, you will know there was a time when desktop/notebook sales were booming and there were many vendors of x86 processors.

As the market matured, the number of players that the industry could support began to shrink until it came down to Intel and AMD, and as we are seeing, AMD is incompetent. Intel has won the PC market and I don't think the ARM ecosystem will do much to change this given how important Windows is to this market.

In servers, there was also once fierce competition, but Intel fought tooth and nail to win that battle too. ARM might make a dent here but given the sheer resources/focus that Intel puts here, I think it's going to be a very uphill battle.

Do you really think that the ragtag players like Cavium Networks, Applied Micro, etc. can really challenge Intel? Qualcomm might be able to but I don't think they will pull off the "A-teams" from their mobile chip biz (their bread and butter) and make the large acquisitions/investments to really fight Intel over the long term despite their claims in the press.

 

[poofyhairguy]

Getting less competitive, though. Economics say that you can't have a bazillion chipmakers all gunning for their piece of a pie that is suddenly not growing so fast. In fact there are only two really viable merchant vendors in mobile: Qualcomm and MediaTek.

That isn't true. Now a lot of Android market phone makers are copying Apple on the path to their own SoC and they are all viable-Samsung, Huawei, and Nvidia.

 

[Nothingness]

That isn't true. Now a lot of Android market phone makers are copying Apple on the path to their own SoC and they are all viable-Samsung, Huawei, and Nvidia.

Nvidia is out of the mobile market.

 

[Arachnotronic]

That isn't true. Now a lot of Android market phone makers are copying Apple on the path to their own SoC and they are all viable-Samsung, Huawei, and Nvidia.

Do you really think the economics of the Android phone market will support every phone maker doing their own SoCs long-term?

p.s. NVIDIA has been out of mobile for a while.

 

[Headfoot]

to understand why things are they way they are you need to take the investor's point of view.

I was under the impression this thread was about A9X geekbench scores and other technical details about the performance of the A9X core. I fail to see what relevance investors and financials have to this thread. Sounds like a topic that could be its own thread.

 

[Arachnotronic]

I was under the impression this thread was about A9X geekbench scores and other technical details about the performance of the A9X core. I fail to see what relevance investors and financials have.

What, you don't think the technical details of the A9X are driven by the economic realities of the industries in which Apple operates?

Understand the economics and your "technical" crystal ball becomes a lot clearer, IMO.

 

[sm625]

So, it looks like it may depend on the type of test? Although, obviously something is happening in that photo enhancement subtest that is out of whack with the other results. Someone should dig into it more.

The WebXPRT photo enhancement test is a very light load on the CPU. I ran the benchmark and set the affinity to the third Core:

It is a very bursty and very light load. When you run Octane with core affinity set to one core, it pegs that core. That is why Speed Shift doesnt really help it. WebXPRT is actually a more realistic benchmark since most typical PC usage doesnt peg cores for the majority of the test like some of these benchmarks do.

 

[dahorns]

The WebXPRT photo enhancement test is a very light load on the CPU. I ran the benchmark and set the affinity to the third Core:

It is a very bursty and very light load. When you run Octane with core affinity set to one core, it pegs that core. That is why Speed Shift doesnt really help it. WebXPRT is actually a more realistic benchmark since most typical PC usage doesnt peg cores for the majority of the test like some of these benchmarks do.

That is what I figured (see my other post on start-stop nature of the bench). It does seem like that test would conform more to typical PC usage. I guess the question becomes "is that 30-100ms perceptible for those tasks?"

I'd think the answer is yes, just based on online gaming experience. But it is something worth pondering.

Anyone find any 3dmark scores for the A9x yet?

 

[witeken]

From a customer perspective yes, but from a business perspective it is downright brilliant and exactly the kind of shareholder-friendly move that Apple should be doing.

Doing the reductio ad absurdum, they should go back to 640kb and charge 100 buck for every doubling. Your ridiculous 16GB can't be justified in any sane way.

Why don't they upgrade the storage to "desktop class"? Take that, $pple $Phone. Same with the ridiculous 1GB memory until 6s, all while you could get 4GB Zenphone 2.

 

[StinkyPinky]

It's amazing how good the A9 and A9X are. Finally, some real competition on the CPU side. Forget AMD, it will be Apple and (hopefully Samsung+Qualcomm etc) that push Intel to actually bother with performance.

Makes you wonder about the A10X. If it shows strong gains again...will they replace their 13" MB Pros and MB Air's with their own CPU's?

Intel must be sweating. If Apple can do it, I have faith that Samsung etc can do it as well. ARM must be loving this.

 

[witeken]

Intel must be sweating.

You are absolutely correct. And GT4 is only next year. RIP Intel. You lost your process lead (while other where still on strained silicon, you had HKMG, while other where still on HKMG, you had FinFET, while other are just on FinFET, you should have had III-V), you lost your Tick-Tock, that thing, that cadance that kept you beating, you lost your short graphics leadership, and now, Intel, you are losing tons of benchmarks on the freaking CPU side not only against Core m, which had its short 2X performance glory and was astonishing for its time, but against 15W SKL-U; you know, that industry you have been dominating for decades, the architecture you have been leading to OoO and other revolutionary architectrures, that clock speed you pushed up to a mighty 8GHz; all that, Intel, even though you are investing more than $11B every year and is more than 20% of reveune, all that you are losing to a [censored] phone company that sells shiny devices for people with IQ << 100?

RIP Intel.

So I guess it's clear I changed my mind. Apple done incredible work, Intel disappointed again with SKL (wow, we got an integrated ISP and a 5th decoder). Apple, take all my money pls (okay, I still won't buy iDevice, but you get the point).

 

[Arachnotronic]

all that you are losing to a [censored] phone company that sells shiny devices for people with IQ << 100?

I had my IQ measured once. It is not <<100

 

[teejee]

You are absolutely correct. And GT4 is only next year. RIP Intel. You lost your process lead (while other where still on strained silicon, you had HKMG, while other where still on HKMG, you had FinFET, while other are just on FinFET, you should have had III-V), you lost your Tick-Tock, that thing, that cadance that kept you beating, you lost your short graphics leadership, and now, Intel, you are losing tons of benchmarks on the freaking CPU side not only against Core m, which had its short 2X performance glory and was astonishing for its time, but against 15W SKL-U; you know, that industry you have been dominating for decades, the architecture you have been leading to OoO and other revolutionary architectrures, that clock speed you pushed up to a mighty 8GHz; all that, Intel, even though you are investing more than $11B every year and is more than 20% of reveune, all that you are losing to a [censored] phone company that sells shiny devices for people with IQ << 100?

RIP Intel.

So I guess it's clear I changed my mind. Apple done incredible work, Intel disappointed again with SKL (wow, we got an integrated ISP and a 5th decoder). Apple, take all my money pls (okay, I still won't buy iDevice, but you get the point).

Didn't expect this post from you. Maybe this post will later be considered as the official turning point for Intel

 

[witeken]

Didn't expect this post from you. Maybe this post will later be considered as the official turning point for Intel

I hope you mean turning point in the good way. I dunno, they should announce something revolutionary.

Like, "we have a real, silicon quantum computer, not the limited quantum annealing thing from DWave" sort of revolutionary. Or some kind of NVM with unlimited endurance and speed like SRAM and NAND density and cost. Or on chip silicon photonics. Or III-V and Ge with Quantum Wells and with secret EUV ("haha we were actually joking about EUV for 7nm, we actually DID something with the 5B we invested in ASML") and 450mm ("yeah, that's why we were spending so many billions of Capex for many years"). Or a new architecture, like that hyped RISC thing ("you thought we delayed 10nm, but that was actually to pull in EUV and 450mm, but as compensation we've got something totally uninteresting: Kabylake with twice the IPC, and double the frequency at 10nm"). Or heck, even near threshold voltage computing. Or a neural chip like IBMs TrueNorth. Or 3D logic.

Guess reality won't be that spectacular (cf. Larrabee) and the lack of revenue growth will just let other companies like TSMC catch up while they reap the benefits of the mobile industry's money.