Thread Title: "Xbox One vs PS4: ESRAM slower than GDDR5, It's Bottleneck"

Actual quote: "Not Much Faster Than GDDR5, Can Never Have Sustained Speed, It's Bottleneck"

Huh?

I'm unaware of any journalistic convention that article titles must exactly replicate quotes from sources.  In fact, they rarely ever do.
ESRAM+DDR is only theoretically faster than GDDR, nobody claims a scenario where it actually is so in a game,
much less specific numbers showing it is faster in a real-world game, or even claiming it to be in a hypothetical game.
That is in line with the article title.

English + Algebra = Spin

To go further, MS also has "Move Engines" on the APU which are dedicated to moving this memory data around thru the ESRAM. 

That also takes room, if not as much as the ESRAM itself.

You can characterize the ESRAM as a bottleneck vs. Sony's solution, because the real world thruut is still lower AND the 32MB window is choking potential workflows that you would like to do with it (and that Sony's GDDR imposes no barrier to), the 32MB window is restricting potential efficiencies.  That can also be characterized as a "development bottleneck", but if the best result a developer can come up with using ESRAM is still lower performance than GDDR, the ESRAM is also imposing a performance bottleneck.  If you want you can say that bottleneck is derived from the DDR, and ESRAM didn't fully compensate DDR's weaknesses, but same result: MS' DDR+ESRAM is a bottleneck.

I believe your are thinking about the bottleneck part to hard and are inventing an opinion that is not sound in theory.  The bottleneck is that the ESRAM at its size and complexity is bottleneck because it either cannot be used at top speed consistantly or memory cannot be moved efficiently to keep it feed.  ESRAM is not the bottleneck its the solution for parts of a game where speed is concerned.  Not all parts of a game where you need to move data at a high data rate and I believe a developer already covered that part.

I assume you're refering to my questioning of the thread title vs the actual quote. The thread title makes one think that the ESRAM in the Xbone is actually slower than the GDDR5 in the PS4. Reading the actual quote, the ' industry insider' actually says the ESRAM is not much faster than the GDDR5. In plain English, he's saying that ESRAM is faster just not by much. Thread title is misleading. It should just say ESRAM causing bottlenecks or ESRAM + DDR3 inferior solution to GDDR5 or something along those lines.

EDIT: I just checked the link and the thread title wasn't even from the link, the poster just made it up.

OK, what is it's 'inferiority'?
Besides displacing GPU cores, and 32mb window constraining rendering approaches, the real world memory thruput is SLOWER.
As I wrote, if MS believes otherwise, and wants to highlight an area where it is superior, it can publish real-game memory thruput stats.

They haven't done so, even for exclusives.
The title is in conformance with that reality. I don't see anything wrong with a thread title reflecting reality.

There is no rule that thread titles MUST ONLY exactly repeat a quote they reference, and cannot be 'made up' by the person creating the thread.
The 1st post contains the accurate quote, up-front for everybody to see and analyze, nothing is being hidden.

Please, this guy isn't an insider. He's just a Sony shill. If you post anti-Microsoft and pro-Sony stuff on NeoGaf long enough you will attain 'insider' status. He's so important that they even misspelled his name.

The PS4 development tools will mature and I would be surprised if all games for it didn't run 1080p/60 in the future. Likewise Xbox One games will run 1080p/60. Both machines are capable enough and once they achieve this, any difference in power will be academic.

The eSRAM has nothing to do with the GDDR5 memory. At all.
You can have eSRAM and GDDR5 memory and it would work fine, they don't share the same die, they are independant of each other.
The Xbox One by it's very nature of having less compute resources also needs less bandwidth.

Basically, Microsoft played it safe, they wanted more Ram for other tasks so the system was more feature rich, without breaking the bank.

Sony was origionally going to go with 4Gb of GDDR5 memory, but decided to take a gamble as new higher density modules were released before the systems launch, so they were then able to double the memory amount for very little cost.
The gamble is that volume production could have been extremely limited as high density GDDR5 doesn't enjoy the same scales of economies that DDR3 currently enjoys and it needs to share whatever volume it has with GPU's on the PC.

It's a gamble that paid off for Sony and well deservedly so.

However... The eSRAM takes up allot of transister space on the APU die, which is not only at a premium, but it forced Microsoft's hand by making them cut back on the graphics processor in order to keep costs down and yields up, otherwise the Xbox one may have cost another $100 extra.
The APU's in the Xbox One and Playstation 4 are already pretty monolithic, so a concession had to be made.

With that in mind, the eSRAM isn't a bottleneck, it's a small "band-aid" solution, the eSRAM isn't going to increase the bandwidth for everything, but a few smaller vital things like render targets which will see massive boons.
Is it ideal as Sony's implementation? Hell no, but apparantly the Xbox One should be identical to the Playstation 4 and isn't allowed to be different in any way shape or form if some peoples comments are a basis to go by, it's good to have differences in a market such as this, otherwise the Wii would never have been made and showed us motion controls.

Basically, the eSRAM doesn't cripple the GPU, it lends its assistance for some tasks, developers also have the choice not to use it at all if it really was a bottleneck and anyone who states otherwise... Are the one's who shouldn't ever be taken seriously again, much like the bloke on the twitter feed.
eSRAM/eDRAM has been used for decades in various forms even on the PC and it has never held/hampered/slowdown/bottlenecked a systems performance.

Great. If MS has real evidence to prove otherwise,
they can publish actual memory bandwidth stats for real games,
ideally cross platform ones where it can be directly compared to PS4.