AMD CPU speculation... and expert conjecture

[mayankleoboy1]

^ +1

 

[cowboy44mag]

For every PS4 and XbawX 720 sold, how many notebooks, office PC's and server chips are sold ?

Spoiler

Intel is not selling ALL of those notebooks, office PCs and server chips. Android is going to take some of those notebook sales as people opt for a Jelly Bean tablet, and AMD is still selling notebooks, office PCs and server chips too. A part of those sales are still going to be going to AMD. ALL the PS4 and Xbox processor profits are going to be going to AMD. Even if AMD had to take a lesser deal and aren't going to make a large profit per console they are still going to make a good profit (maybe not as much as they would like, but there will still be a good profit margin) and there are millions of consoles that sell in the US alone, let alone the world market. Sony and Microsoft are also not as concerned with the profit they get from the console sales as they make the vast majority of their profit trough software sales. The Xbox 360 and PS3 console sales didn't even make a profit till the consoles were on the market for over three years (but the hardware suppliers still made good profits). The very first day of release there will be millions of PS4 and Xboxs sold.

..

Compare the profit margin of selling a server chip, and the associated maintainance contract, to the tiny, one-off payment per chip.

I know PC lovers don't like to admit it, but console sales far outweigh gamer PC sales, which is why studios produce games for console systems and port them for PC, not the other way around. AMD will lead the sales of the gaming industry, which is no small niche, and will be a large contributing factor as to why there aren't going to be bought up by Intel or just go belly up

Who said anything about gaming PC ? They are about 1% of the non-server x86 market. "AMD leading the game industry" will do what to AMD's bottom line ?

As to if the Sony and Microsoft contracts were a bitter pill to swallow because AMD didn't have a choice, or if it was strategy is irrelevant as it means the same thing in the end. Game studios will still have to contend with AMD processors and will therefore have to produce mulit-threaded games (utilizing 4-6 cores) that will on the software end favor AMD. I'm sure Intel will still have a slew of benchmark software that shows them way way way out in front of AMD, but if your pricey Haswell isn't gaming better or as well as a less expensive Steamroller alternative I can see a lot of Intel fans getting really mad

If you are talking about CPU performance for games, then Intel has 4 and 4+4 core variants as well. And those perform better than 8 core PD in most games. And Haswell will do better than a SR chip.

You can't possibly say "Haswell will do better than a SR chip" unless you are the first person is history to figure out time travel. So do we got Dr Who on this topic then or what? Neither chip has been released, there are no reviews of the actual hardware and there is no way of knowing which one will be better. It is Intel fanboyism and ego at its most extreme.

I personally love history, do you? If there is one thing that history shows us its you can't possibly predict the future. If you were to ask a military strategist in 1776 if the colonies of the new world could ever best the British Empire how hard do you think the know-it-alls would have laughed? Yet we won, and became the most powerful military might the world has ever known. You can't possibly predict accurately or know how Steamroller or Haswell will perform. Steamroller may be great, it may only be marginally better than Piledriver. Haswell could be so great that NASA uses it to colonize Mars, or it could be Intel's version of Bulldozer and be a big fudd, no one knows.

 

[Cazalan]

For every PS4 and XbawX 720 sold, how many notebooks, office PC's and server chips are sold ?

The Gartner estimate is 352 million PCs sold in 2012.

http://www.gartner.com/newsroom/id/2301715

XBox 360 sales were 72 million total (since launch till 2013)

http://www.gamespot.com/news/77-million-xbox-360s-sold-6407243

PS3 sales at 77 million total (since launch till 2013)

http://www.gamesindustry.biz/articles/2013-01-09-idc-game-consoles-discs-to-remain-revenue-mainstays-for-years-to-come

PC sales far outsell the consoles combined sales.

 

[sarinaide]

Dev costs have skyrocketed, on the old systems
Thats going to change.
They eked out about as much as they could on the old systems, which even today, some games bring PCs to their knees.
They have 8x available on the new consoles, plus ease of dev costs.


I think therell be abuse and laziness racing to release of newer titles/games, but thats just it isnt it?
Everyone will be able to do this, its just, I have a feeling everyone wont

Thats what I am on about, its the nonchalant attitude towards the PS4/XBox and the fact that the APU brings hardware and software into uniformity for untapped resources. By owning all the platforms AMD will now have optimizations as developers will want to achieve the full potential of the PS4/XBox which is said to be quite limitless relative to old cell core based systems. There is now a parity between hardware and software maximising potential that the hardware itself will become less pertinent.

I don't quite get the baseline those slides were refering to, they failed to give scores which is odd in its own right but despite that at least Intel has a iGPU solution albeit only at the Iris level. Its nearly a year now since Trinity and Intel have been targeting it as the criterion to beat, AMD have themselves stated that its not a very high benchmark as they term Trinity as a infant in iGPU performance. The real question with Iris and the rumors I am hearing as to costing is that a buyer of any GT3 part will need to make very certain that integrated is the way they are going as the pricing is not very cheap, compared to the 4770K its a bit of a gut check.

TH showed 3DMark 11 Graphic score of 652 for the HD 4000 in the i7-3770k. That would put the i7-4770R at 1956.

http://www.tomshardware.com/reviews/ivy-bridge-benchmark-core-i7-3770k,3181-4.html

The A10-5800K gets a score of 1183.

http://www.tomshardware.com/reviews/a10-5800k-a8-5600k-a6-5400k,3224-8.html

That would give a 65% advantage to the GT3e. Of course that is just a synthetic benchmark and may have vastly different results in a real game. However that one of the same benchmark AMD cites to talk about performance gains.

Thats 3x the score and faster than a HD6670/GTX260, I am calling BS story. When the HD4600 leaks came out the HD4000 scored around 762 marks, the prototype i5 with 4600 scored 732 and the 4770K rumored around 800 marks. Since the composition is not scaled that much higher I can't see HD5000+ achieving more than the 7660D's 1300-1500 marks Possibly around the 1100mark range.

I am also getting tired of that muppet from Anandtech saying GT3 is a GTX650M beater it is lunacy of the highest level. For all I know those charts could be more liberal, never has 3x performance ever remotely been achieved and there is no inclination to believe it will now unless the clockspeed is in overdrive.

 

[cowboy44mag]

For every PS4 and XbawX 720 sold, how many notebooks, office PC's and server chips are sold ?

The Gartner estimate is 352 million PCs sold in 2012.

http://www.gartner.com/newsroom/id/2301715

XBox 360 sales were 72 million total (since launch till 2013)

http://www.gamespot.com/news/77-million-xbox-360s-sold-6407243

PS3 sales at 77 million total (since launch till 2013)

http://www.gamesindustry.biz/articles/2013-01-09-idc-game-consoles-discs-to-remain-revenue-mainstays-for-years-to-come

PC sales far outsell the consoles combined sales.

Agreed, what I was talking about with consoles outselling PCs is the gaming market. One more thing to take into consideration though is how far has ARM Android come in the last year? How many people who bought PCs last year are going to get a new Android tablet and figure they no loner need the shackles of a bulky desktop or laptop when the hand held tablet can do near everything the conventional PC can? Most "light" users want to surf the internet, Tweet, use Face Book, Skype, do light word processing, ect. A tablet can do all those things and take pictures too. Really for home PCs the only advantage of a real laptop or better yet desktop is gaming. A tablet simply can't compete in the gaming world. But for people who don't game on their computers I think the trend towards smartphones and tablets has already cut into PC sales and will only get bigger going into 2014. People who opted for a tablet over desktop who want to play video games are going to turn even more to the console systems.

It sucks, as I personally love my desktop computer, and would surely break a tablet within a week, but it is the current trend. Thankfully people who still want to game on their desktop PCs make up a good enough market to keep desktops around for a little while longer at least.

 

[8350rocks]

---------------------------------------------------------------------------------------------------------------------------------
Just for the team green crowd

https://amd.box.com/3dmark

Who was talking about nvidia was superior because of bad drivers? Whoever it was, there's you newest version of catalyst...looks like a winner to me.

 

[cowboy44mag]

I have to go and get some work done now, but I just want to send out a thank you to everyone who has posted in this discussion. I hope that I have not outright offended anyone, as it was not my intent, its no secret that I don't like Intel or the fanboy rhetoric, but I got to give Intel credit they make fine chips. I just like all the posts, thoughts and ideas that show that AMD also makes some fine chips!! Thank you all for the great posts, links, information and ideas. This is one of my all time favorite discussions, please keep all the great thoughts coming!!

 

[JAYDEEJOHN]

First of all, having all the market of 144 million spread over x amount of years, where most sold early on, lets say 20 million a year early on, and this is exclusive, and both gpu and cpu.
Looking at AMDs marketshare currently, 352 million, generously giving AMD 20%, thats 70 million, so approximately 30% of their sales, like whats already been said by AMD.

As far as lazy devs goes, if the indies dont get shut out, it would be a crime some indie comes out and smashes whats the norm for quality against the big houses wouldn't it?

 

[mayankleoboy1]

You can't possibly say "Haswell will do better than a SR chip" unless you are the first person is history to figure out time travel. So do we got Dr Who on this topic then or what? Neither chip has been released, there are no reviews of the actual hardware and there is no way of knowing which one will be better. It is Intel fanboyism and ego at its most extreme.

Well Tomshardware leaks showed the perf of a 4770K. And the most realistic expectations from SR is ~15-20% over PD.
Based on that i would say that Haswell > SR in CPU.

And neither AMD nor intel have much of a presence in the Tablet market. So the trend towards mobility hurts both of them.

 

[sarinaide]

---------------------------------------------------------------------------------------------------------------------------------
Just for the team green crowd

https://amd.box.com/3dmark

Who was talking about nvidia was superior because of bad drivers? Whoever it was, there's you newest version of catalyst...looks like a winner to me.

There were people lingering on AMD's past but in theory AMD sandbagged GCN for long enough for Nvidia to pander there parts as hero parts only to release 7-8 driver updates in a year culminating in Nvidia having no recommendable part anymore; The GTX670 is probably it but the 7870XT delivers within 5-10% of its performance at $150 less, the 650ti Boost is Slower than the 7850 and only now marginally faster than the 7790 all the while the Nvidia card sucks more power. If nvidia drivers were so amazing then why has their been no real progression. This round of graphics cards was a convincing AMD success on performance and value. New drivers improve CFX substantially and AMD has already said a new batch of drivers will be released in succession to there goes the microstutter arguement.

 

[cowboy44mag]

You can't possibly say "Haswell will do better than a SR chip" unless you are the first person is history to figure out time travel. So do we got Dr Who on this topic then or what? Neither chip has been released, there are no reviews of the actual hardware and there is no way of knowing which one will be better. It is Intel fanboyism and ego at its most extreme.

Well Tomshardware leaks showed the perf of a 4770K. And the most realistic expectations from SR is ~15-20% over PD.
Based on that i would say that Haswell > SR in CPU.

And neither AMD nor intel have much of a presence in the Tablet market. So the trend towards mobility hurts both of them.

Just have a minute here... "Tomshardware leaks showed the perf of a 4770K" more speculation made by experts about what may be in the future. Bulldozer was also supposed to be awesome before its release, it was supposed to bulldoze Sandy Bridge under.... didn't happen, but experts were speculating it would. Until they are actually released there is no way to know which will be better. It may be that the 4770K will perform well in certain niches (especially programs where single core execution are important) but will give room or be beaten by Steamroller in other niches (new games ect) where mulit-threading is most important.

"And neither AMD nor intel have much of a presence in the Tablet market. So the trend towards mobility hurts both of them" Indeed, however what I was trying to point out is the people who opted for mobility in a tablet may still want to video game, have a child or grandchild who wants to video game. Those people are going to go for a console system, even if they use to game on their PC. The trend towards Android tablets is going to drive the sales of console systems up as you can't game on a tablet (not the high end games on PC and console). Most people will upgrade a tablet every two or so years but will consider a console (life span ~7 years) a good gaming solution.

 

[anxiousinfusion]

so, is this the real reason amd's stocks are up suddenly?
"Intel Could Buy AMD as x86 Microprocessors Lose Share to ARM – Analysts.
AMD’s Shares Are Up on Intel Buyout Rumour"
http://www.xbitlabs.com/news/cpu/display/20130501223936_Intel_Could_Buy_AMD_as_x86_Microprocessors_Lose_Share_to_ARM_Analysts.html
i was wondering whom RR was polishing amd up for....

AMD Establishes Semi-Custom Business Unit to Create Tailored Products with Customer-Specific IP.
http://www.xbitlabs.com/news/other/display/20130501231534_AMD_Establishes_Semi_Custom_Business_Unit_to_Create_Tailored_Products_with_Customer_Specific_IP.html

*picks up a bowl of popcorn and waits for amd fanboys to implode*

edit: personally, i think intel should buy nvidia and appoint the ceo to it's gfx (consumer, server and professional) division as some kind of cto. but i guess amd is much cheaper and they don't have to pay glofo (timing is near-perfect).... :lol:

I'm just entertaining the idea of an APU; 22nm, Haswell cores + Radeon GPU. It would never happen but this would be a theoretically killer product.

I remember when i first got into all of this i was reading console vs PC articles and Desktop is dieing and Intel will buy Amd or Amd is doomed articles all the time and this was over 7 years ago now i'm reading them again. Lol Maybe technology changes but the Questionable writers don't.

The technology changes, the people don't.

Also, when HSA is introduced with Kaveri, how will hybrid crossfire with a dGPU affect HSA in 3D applications? Will it completely nullify the gains? I can only imagine the crossfired card would only force the APU to handle graphics loads traditionally.

 

[8350rocks]

Also, when HSA is introduced with Kaveri, how will hybrid crossfire with a dGPU affect HSA in 3D applications? Will it completely nullify the gains? I can only imagine the crossfired card would only force the APU to handle graphics loads traditionally.

I should think that when the drivers come out and you can pair it with higher end HD7XXX series cards, you'd nearly be a fool not to use a CF setup, which would only magnify the gains from HSA.

 

[Cazalan]

Thats 3x the score and faster than a HD6670/GTX260, I am calling BS story. When the HD4600 leaks came out the HD4000 scored around 762 marks, the prototype i5 with 4600 scored 732 and the 4770K rumored around 800 marks. Since the composition is not scaled that much higher I can't see HD5000+ achieving more than the 7660D's 1300-1500 marks Possibly around the 1100mark range.

I am also getting tired of that muppet from Anandtech saying GT3 is a GTX650M beater it is lunacy of the highest level. For all I know those charts could be more liberal, never has 3x performance ever remotely been achieved and there is no inclination to believe it will now unless the clockspeed is in overdrive.

Well you have to read the fine print. It was a single benchmark from the 3DMark 11 suite. Until we know more about the inner workings of Crystallwell (64/128MB eDRAM) its not a stretch. It's like the difference between a benchmark that fits entirely in the L3 cache vs one that is too large and needs to keep going to main memory. The bandwidth and latency difference is significant.

I have no interest in buying Crystallwell but I am definitely curious about the performance tradeoffs they can get by having a large embedded memory.

If you go way back the difference between a Pentium and Celeron was the Celeron had no L2 cache. The performance difference was like night and day. Now a 4th layer of cache is being introduced and some operations will definitely see a big improvement with that. IBM has been using eDRAM in their servers chips for years and they wouldn't do that without a good reason.

 

[juanrga]

As a respected reviewer told me recently, technology has come on leaps and bounds, the potential of the APU is astounding and the new generation consoles are largely untapped in its true performance. We are at the point where hardware to hardware is not the defining feature but how now software interlinks. Software writers are 10 years behind, while some are deciding to change it is now software playing catch up.

---------------------------------------------------------------------------------------------------------------------------------
Just for the team green crowd

https://amd.box.com/3dmark

In one sense, Bulldozer/Piledriver suffered from software not using all their real potential. It is just now when we start to see recent games such as Crysis 3 showing how a FX-8350 can outperform an i7-3770k.

Just curious what the respected reviewer told you was about the potential of the APU or of the APU-HSA? Because I agree about the latter being a kind of quantum leap.

 

[juanrga]

I have poor memory. IIRC, the perf improvement of Richland GPU is ,what, 15% over Trinity ?
Which means GT3e is ~60% the perf of Richland.

 

[juanrga]

Does anyone understand the main reason they have this many cores is for background tasks? I even heard a rumor that PS4 and the 720 will use 2 cores just for the OS.

I could see it for the Infinity; Windows itself would need a core, and it looks like the CPU that manages the Kinnect interface has moved off the Kinnect proper and onto the main console, so that would require a separate CPU due to all the I/O the Kinnect would be pumping out. Also remember the Infinity supports multiple apps at once, so that also could eat CPU cores. I REALLY want to know how much the console is GUARANTEED for developer use for a single app at any one time due to these issues. I honestly wouldn't be shocked if you can only guarantee 4-5 cores for the main foreground application.

For the PS4, if Sony keeps a slim OS like it has now, one core is plenty. I really don't see them implementing a heavy OS...

Eight-cores are accessible to game developers and they will be programming for eight unlike what happened with previous consoles where they programmed for single-core essentially

http://www.eurogamer.net/articles/digitalfoundry-future-proofing-your-pc-for-next-gen

 

[JAYDEEJOHN]

Now you could dedicate a core or two to the OS, but it doesnt need it, and if it doesnt fully use it/them, it would be a waste of resources.
Others pointed out, the OS determines everything, no need for a dedicated core, unless its the ARM core, which could be as powerful as those old cores or moreso, and switching to borrow another if needed, when not in standby/low power etc using HSA and hUMA

 

[juanrga]

The issue is this: The PS3 especially (less so for the 360) needed to be coded at a VERY low level to get any real performance and keep the memory footprint around ~200kb or so. So you have games with a lot of hardcoded thread logic, memory management, and the like. This makes it hard to reuse code for the PC version, driving up costs and delaying development (since you need ot basically re-write your memory and GPU interfaces, just for starters).

This gen, boom, you have more hardware then you know what to do with. And its a PC like architecture. The result is likely to be that all those optimizations on consoles to get the most out of them will stop being used in order to make it trivial to re-use large parts of the code base between both consoles and the PC. So over the long term, you won't see as much advancement this console generation, as most of the hardware won't be used at max capacity. This will, also over the long term, stagnate the PC, as you will see the consoles bloated specs actually hold back the PC.

In short: I argue the PC like hardware on consoles is going to stagnate PC development for the next decade, because there will be no incentive to spend the time and money to really get the most out of the new consoles.

That "VERY low level" coding was needed to compete with 10x more powerful PCs.

First generation of next games will be probably fast ports from PC games/codes. But how PC hardware advances, console programmers will start to use optimization to compete with much faster PCs.

The advantages now are that (i) the PC-like hardware simplifies optimizations a lot of when compared with previous consoles and (ii) some of those optimizations could be reused in PCs with 'compatible' hardware. E.g. in PCs with AMD APUs.

 

[juanrga]

For every PS4 and XbawX 720 sold, how many notebooks, office PC's and server chips are sold ?

Intel is not selling ALL of those notebooks, office PCs and server chips. Android is going to take some of those notebook sales as people opt for a Jelly Bean tablet, and AMD is still selling notebooks, office PCs and server chips too. A part of those sales are still going to be going to AMD. ALL the PS4 and Xbox processor profits are going to be going to AMD. Even if AMD had to take a lesser deal and aren't going to make a large profit per console they are still going to make a good profit (maybe not as much as they would like, but there will still be a good profit margin) and there are millions of consoles that sell in the US alone, let alone the world market. Sony and Microsoft are also not as concerned with the profit they get from the console sales as they make the vast majority of their profit trough software sales. The Xbox 360 and PS3 console sales didn't even make a profit till the consoles were on the market for over three years (but the hardware suppliers still made good profits). The very first day of release there will be millions of PS4 and Xboxs sold. I know PC lovers don't like to admit it, but console sales far outweigh gamer PC sales, which is why studios produce games for console systems and port them for PC, not the other way around. AMD will lead the sales of the gaming industry, which is no small niche, and will be a large contributing factor as to why there aren't going to be bought up by Intel or just go belly up.

As to if the Sony and Microsoft contracts were a bitter pill to swallow because AMD didn't have a choice, or if it was strategy is irrelevant as it means the same thing in the end. Game studios will still have to contend with AMD processors and will therefore have to produce mulit-threaded games (utilizing 4-6 cores) that will on the software end favor AMD. I'm sure Intel will still have a slew of benchmark software that shows them way way way out in front of AMD, but if your pricey Haswell isn't gaming better or as well as a less expensive Steamroller alternative I can see a lot of Intel fans getting really mad.

Yes. A game as Skyrim sold about 4x more copies for consoles than for PCs.

 

[juanrga]

If you are talking about CPU performance for games, then Intel has 4 and 4+4 core variants as well. And those perform better than 8 core PD in most games.

Do you mean if you count the vast amount of older games, which were single-threaded, and the recent one being dual or quad. But this is not true for the next generation of six- octa-treaded games

You seem unaware, but all triple-A game developers poll by Eurogamer selected the AMD eight-core FX-8350 chip as the better choice for a future gaming rig.

And Haswell will do better than a SR chip.

Do you mean some wait so? Other wait the contrary

http://news.softpedia.com/news/AMD-s-Steamroller-To-Be-Faster-than-Intel-Haswell-289980.shtml

http://www.overclock.net/t/1356581/amd-steamroller-the-secret-weapon-that-could-decimate-intel

 

[gamerk316]

So technology that opens up avenues into more efficient codes and software implementation, and to be frank this technology is screaming at you writers to untap plenty potential. There have been a number of reports from AMD and game writers which are of the opinion that AMD's open path to HSA is simple to code in terms of not being foreign but exciting going forward. I don't understand why you continually seem to reject this idea.

HSA itself is a good idea, but understand that it clashes with the idea of a GPU with its own cache. So something has to give on that front. Its possible that Intel/AMD eventually make discrete GPU's uncessesary, or GPU's lose the cache and access main memory. But until then, HSA is only valid for integrated GPU's. And from a coding perspective? No chance necessary, as its all hardware driven.

You also have major downsides: Currently, GPU data is independent from the OS (CPU) Address Space; eg, only about 512MB is reserved at any one time, despite the GPU's large (2GB) cache. So its quite possible to have a memory footprint of 3GB or so, while still having large texture caches. One address space though, and guess what? 4GB limit, period. Think textures are going to cut into that quite a bit? [Granted, this is a non-issue once 64-bit becomes standard, but don't be shocked if this causes 8GB+ systems to become mandatory]

We are now moving on to new Unreal engines and Frostbite engines that are legitimately using 4-8 cores effectively and the scaling is showing.

It doesn't help FPS though. Which is the entire point, isn't it?

I've actually started a thread analysis for BF3 using GPUView and PerfMon. Preliminary stuff right now, but it looks like three threads do the majority of the work; two render threads (DX11) and the main game thread. A handful of others do minor background work, and a bunch do <1% workloads. So you have a game that scales to about 3-4 cores, and after that point, the number of cores won't improve performance, just lower the overall usage statistics somewhat.

Cryteks own engine was capable of using more than 4 cores effectively and this leveraged performance with the 8350 capable of drawing upto the 3770.

But you missed the discussion on how they did that; it wasn't any special property of the game engine, they instead moved work from the GPU onto the CPU. Which sounds all nice and dandy, until you remember two facts:

1: Parallel workloads execute faster on the GPU
2: GPU performance is increasing several times faster then GPU performance.

On non-GPU bottlenecked systems, what Crytek did for Crysis 3 will end up reducing performance due to a CPU bottleneck. Hence why I consider what they did idiotic.

 

[gamerk316]

and the fact that the APU brings hardware and software into uniformity for untapped resources.

Guess what? No SW changes for HSA. Memory Addressing is not the domain of software (from a developer perspective at least), but the OS.

 

[juanrga]

I am also getting tired of that muppet from Anandtech saying GT3 is a GTX650M beater it is lunacy of the highest level.

They can go poor. Now Intel is hyped as the new GPU company that competes with "performance mainstream"...

Intel’s performance target for the highest end configuration (GT3e) is designed to go up against NVIDIA’s GeForce GT 650M, a performance target it will hit and miss depending on the benchmark.

Regardless of whether or not it wins every benchmark against the GT 650M, the fact that an Intel made GPU can be talked about in the same sentence as a performance mainstream part from NVIDIA is a big step forward. Under no circumstances could Intel compete with NVIDIA on performance and still do so under the Intel HD Graphics brand. Haswell is the beginning of a new era for Intel. The company is no longer a CPU company forced into graphics, but with Haswell Intel begins its life as a GPU company as well. As a GPU company, Intel needs a strong GPU brand. AMD has Radeon, NVIDIA has GeForce, and now Intel has Iris.

 

[gamerk316]

Eight-cores are accessible to game developers and they will be programming for eight unlike what happened with previous consoles where they programmed for single-core essentially

http://www.eurogamer.net/articles/digitalfoundry-future-proofing-your-pc-for-next-gen

Uhh...no. Your second part of the statement is simply BS.

There seems to be a general lack of understanding of the differences in coding to a dedicated piece of hardware with a very light OS, and a general purpose PC with Windows.

Now you could dedicate a core or two to the OS, but it doesnt need it, and if it doesnt fully use it/them, it would be a waste of resources.

Here's the thinking on that one: By putting the OS on a single core, you know NOTHING will be running on the other seven. Now, if thread management/core loading is still handled is useland (rather then by the OS, as in Windows), then your program is essentially deterministic: At any period of time, I know EXACTLY what is running, where every piece of data is stored in memory, and whatnot. This allows for VERY powerful performance optimizations.

If the thread scheduling is up to the OS though, then you can not guarantee what threads are being run at any point in time. You have a non-deterministic program, like PCs have. This basically forbids any form of assembly optimizations that make any assumptions in regards to the contents of any other threads. Granted, its a workload off the developer, but you lose some performance as a result.

So the question becomes, would a loss of one core be offset by optimizations that could be made for the other seven?

Others pointed out, the OS determines everything, no need for a dedicated core, unless its the ARM core, which could be as powerful as those old cores or moreso, and switching to borrow another if needed, when not in standby/low power etc using HSA and hUMA

Remember, you are still thinking in PC/Windows terms. Console OS's up until now have been VERY light.