• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Ars Technica: Penello's XB1 Numbers, Ars "Sniff Test" Analysis

jaypah

Member
Lol, this is going to be the exact same thread all over again but with the twist that the OP cites the first thread. Inception.
 

Krakn3Dfx

Member
new theory: Albert got trolled by his coworker

Penello: "Ok dude, are you SURE this stuff is accurate? Those cats on GAF are pretty rough"

Fellow: "Dude, it's totally accurate, go for it" trollface.jpg

This, Jen, is The Internet
internet.jpg
 

ascii42

Member
I'm waiting for Albert now, maybe he will explain what he meant by that.

I'm guessing he's saying that because there are 8 cores, each of which got a 6% boost, it adds up. I suppose that's true, but it's still just a 6% performance increase.
 

alr1ght

bish gets all the credit :)
Lol, this is going to be the exact same thread all over again but with the twist that the OP cites the first thread. Inception.
It's a well known site calling bullshit on his claims, not just random gaf posters.
 

Raist

Banned
Well there's a surprise.

Looking forward to a reply. Buckling in.

Don't hold your breath. When called out on the BS and asked for explanations, he went into "Well I give you guys proofs straight from our elite engineers but you just think I lie. WE MADE DIRECT X!" and "it's pointless for me to post more about the topic, you don't trust me. You'll see at launch anyway" mode.
 

jaypah

Member
It's a well known site calling bullshit on his claims, not just random gaf posters.

But the responses will be the same since everybody here already took their position. I'm not saying it doesn't deserve to be a thread, just that I expect it to be very similar to the other one.
 

McHuj

Member
Penello: "18 CUs [compute units] vs. 12 CUs =/= 50% more performance. Multi-core processors have inherent inefficiency with more CUs, so it's simply incorrect to say 50% more GPU."

Ars: "The entire point of GPU workloads is that they scale basically perfectly, so 50% more cores is in fact 50% faster."

Except GPU benchmarks usually don't support this view. Even if you compare GPU's in the same line, performance never scales perfectly. For example, look at Anandbench for a comparison of the 7870 vs 7970. The Tflop difference is +48% in favor of the 7970, yet in benchmarks it performs on average ~30% better.
 

ElTorro

I wanted to dominate the living room. Then I took an ESRAM in the knee.
It's a well known site calling bullshit on his claims, not just random gaf posters.

Nice to get some backup from a "respectable" source since many people dismiss statements from TechGAF because of the anonymity. Nevertheless, if a majority of TechGAF starts to bring the same counter arguments independently, then that in itself is evidence that things are fishy.
 

Krakn3Dfx

Member
Don't hold your breath. When called out on the BS and asked for explanations, he went into "Well I give you guys proofs straight from our elite engineers but you just think I lie. WE MADE DIRECT X!" and "it's pointless for me to post more about the topic, you don't trust me. You'll see at launch anyway" mode.

If only I knew how to paste Penello's head into this gif properly.
tumblr_mhm984aR5W1s1wblho1_400.gif
 
I'm on mobile so I can't give a link but ERP on B3D (Mod and former first party Sony dev) said that you have have a performance decrease per CU the more you have. And the rest is basically "we don't know better".
Meh article

Was it this?

FWIW my expectation is that PS4 ought to have a performance advantage, but I wouldn't expect it to reflect the difference in CU counts. CU's are MASSIVELY underutilized on vertex heavy workloads and plenty of the frame will be ROP or bandwidth limited.
There are just too many little things that can significantly impact performance, there were times with early firmwares on PS4 where seemingly innocuous changes would affect performance by as much as 10%.

The eSRAM will certainly provide an advantage under some circumstances, and I'm interested if the ROP difference will end up being a factor, or the lower eSRAM latency will end up nullifying it.

For that matter I could imagine the graphics API having a significant effect, the CPU single threaded performance isn't great on these machines and a poorly conceived implementation, could hurt games across the board. Sony having "lower level access" here isn't necessarily a win.

Hell I could imagine cases where CPU limited games demonstrate an advantage on XB1.
http://beyond3d.com/showpost.php?p=1782673&postcount=6777
 

DeviantBoi

Member
For some reason I just feel that when this guy and Major Nelson start talking hardware, they're just parroting what the hardware team tells them.

"Yes, it's just as powerful as the PS4," MS' hardware team tells them.

So that's what they tell people.

When Cerny speaks, however, I know that he designed the damn thing. When he talks numbers, I know that he knows what they mean, even though I might not fully understand him. If he says the PS4 is more powerful, I sure as hell believe him.
 

velociraptor

Junior Member
Except GPU benchmarks usually don't support this view. Even if you compare GPU's in the same line, performance never scales perfectly. For example, look at Anandbench for a comparison of the 7870 vs 7970. The Tflop difference is +48% in favor of the 7970, yet in benchmarks it performs on average ~30% better.
Yes it only performs 30% better. What a worthless improvement!
 

Flatline

Banned
Penello: "We have more memory bandwidth. 176gb/sec is peak on paper for GDDR5. Our peak on paper is 272gb/sec. (68gb/sec DDR3 + 204gb/sec on ESRAM). ESRAM can do read/write cycles simultaneously so I see this number mis-quoted."

Ars: "Just adding up bandwidth numbers is idiotic and meaningless. While the Xbox One's ESRAM is a little faster, we don't know how it's used, and the PS4's GDDR5 is obviously a lot bigger."


It should also be added that the 204gb/sec number is fake too since they're adding reads and writes in a single bus ESRAM. The ESRAM is 109gb/sec so in fact slower.
 

QaaQer

Member
For some reason I just feel that when this guy and Major Nelson start talking hardware, they're just parroting what the hardware team tells them.

"Yes, it's just as powerful as the PS4," MS' hardware team tells them.

So that's what they tell people.

When Cerny speaks, however, I know that he designed the damn thing. When he talks numbers, I know that he knows what they mean, even though I might not fully understand him. If he says the PS4 is more powerful, I sure as hell believe him.

And the fact that he has zero education in marketing and PR.
 

le.phat

Member
WE GET IT. PS4 IS FASTER. What in fuck's name is the point of this all ? Albert WORKS for Microsoft, its his job to make the difference seem as marginal as possible. He will never come out and say "chucks guys, you got me". Xbone fanboys will try to downplay the difference tooth and nail. Do we have to do this each and every day ?
 
Another one of these threads I see.

It would have been nice of Ars to actually respond with something insightful instead of just spouting off the cuff remarks.

Whatever pleases their readers, I guess.
 

Coconut

Banned
So pretty much no one knows anything about the system until someone outside of Microsoft gets there hands on a fresh off the line box. Great article...
 

Freki

Member
WE GET IT. PS4 IS FASTER. What in fuck's name is the point of this all ? Albert WORKS for Microsoft, its his job to make the difference seem as marginal as possible. He will never come out and say "chucks guys, you got me". Xbone fanboys will try to downplay the difference tooth and nail. Do we have to do this each and every day ?

For some reason some people don't like FUD...
 

alr1ght

bish gets all the credit :)
WE GET IT. PS4 IS FASTER. What in fuck's name is the point of this all ? Albert WORKS for Microsoft, its his job to make the difference seem as marginal as possible. He will never come out and say "chucks guys, you got me". Xbone fanboys will try to downplay the difference tooth and nail. Do we have to do this each and every day ?

You can't really spin math. Being honest goes a long way.
 
Still waiting for Arthur Gies to pop up at the first LensOfTruth/DigitalFoundry multiplat face off with XB1 winning calling out everyone doubting the power of the secret sauce.
 

Dragon

Banned
WE GET IT. PS4 IS FASTER. What in fuck's name is the point of this all ? Albert WORKS for Microsoft, its his job to make the difference seem as marginal as possible. He will never come out and say "chucks guys, you got me". Xbone fanboys will try to downplay the difference tooth and nail. Do we have to do this each and every day ?

1. Don't complain about the existence of threads.
2. Albert is sharing inaccurate information.
3. Calm down, life will be okay.

Another one of these threads I see.

It would have been nice of Ars to actually respond with something insightful instead of just off the cuff remarks.

It would be nice for people to stop spreading misinformation too. But that's life!
 

astraycat

Member
Was it this?

FWIW my expectation is that PS4 ought to have a performance advantage, but I wouldn't expect it to reflect the difference in CU counts. CU's are MASSIVELY underutilized on vertex heavy workloads and plenty of the frame will be ROP or bandwidth limited.
There are just too many little things that can significantly impact performance, there were times with early firmwares on PS4 where seemingly innocuous changes would affect performance by as much as 10%.

The eSRAM will certainly provide an advantage under some circumstances, and I'm interested if the ROP difference will end up being a factor, or the lower eSRAM latency will end up nullifying it.

For that matter I could imagine the graphics API having a significant effect, the CPU single threaded performance isn't great on these machines and a poorly conceived implementation, could hurt games across the board. Sony having "lower level access" here isn't necessarily a win.

Hell I could imagine cases where CPU limited games demonstrate an advantage on XB1.
http://beyond3d.com/showpost.php?p=1782673&postcount=6777

The funny thing about him talking about the diminishing returns of the CUs is that he mentions that they are ROP/bandwidth limited. In these cases, the PS4 has twice the ROPs, and 3x the bandwidth to main memory -- the performance difference would much more heavily favor the PS4 in these cases.
 
WE GET IT. PS4 IS FASTER. What in fuck's name is the point of this all ? Albert WORKS for Microsoft, its his job to make the difference seem as marginal as possible. He will never come out and say "chucks guys, you got me". Xbone fanboys will try to downplay the difference tooth and nail. Do we have to do this each and every day ?

The problem is he comes here everytime and says that, although it's already been discussed to death these figures are not true. If he would just lay it to rest instead of trying to convince everyone that the difference is marginal it would be alright.
But he's spreading this bullshit over and over again. That's what makes it annoying.
 

cchum

Member
You'd think that Microsoft would like to have large amounts of highly skilled knowledgeable people working for them ... that understand the basics of their technology lol.

They do. They just don't talk to the public. Penello is doing HIS job for better or worse aka making/trying to make Microsoft look better.
 

Averon

Member
WE GET IT. PS4 IS FASTER. What in fuck's name is the point of this all ? Albert WORKS for Microsoft, its his job to make the difference seem as marginal as possible. He will never come out and say "chucks guys, you got me". Xbone fanboys will try to downplay the difference tooth and nail. Do we have to do this each and every day ?

This flare up in the PS4 vs. XBO tech specs started again when MS PR tried to damage control the power difference between the PS4 and XBO. We're talking about it because MS made it an issue, and continue to make it an issue every time people like Albert comes here and spew BS like no one here has the technical know-how to discern whether the PS4 really is more capable than the XBO. If they'd stop trying vainly to spin the numbers, I'd say most of this conversation would go away.
 
All he had to do was say no, the X1 does not have an additional discrete GPU (LOL) but nope, he had to throw in the ps4 comparison at the end.
 

EGM1966

Member
WE GET IT. PS4 IS FASTER. What in fuck's name is the point of this all ? Albert WORKS for Microsoft, its his job to make the difference seem as marginal as possible. He will never come out and say "chucks guys, you got me". Xbone fanboys will try to downplay the difference tooth and nail. Do we have to do this each and every day ?

  • This is a gaming site
  • The MS guy posted on this site directly with what is being challenged as Spin and FUD
  • The launch of PS4/XB1 is big news particularly given the drama around price point difference, projected performance differences and the potential for a market swing

Yes we're going to do this every day.
 

Faustek

Member
To quote myself.


Not everyone will understand all that but it was a damned good post.
I just tell my niece this when she wanted an explanation.
You run faster than I do right? But I'm stronger right?
So what goes faster, you running back and forth from the car and taking two shopping bags at the time or me taking all eight at once?
And she usually gets that better than all technical explanations but she is getting there, can't expect a 13 year old to understand everything, heck even I have to Google up bits of info nearly every damn day since I forgot something....it sucks getting old folks.
 
WE GET IT. PS4 IS FASTER. What in fuck's name is the point of this all ? Albert WORKS for Microsoft, its his job to make the difference seem as marginal as possible. He will never come out and say "chucks guys, you got me". Xbone fanboys will try to downplay the difference tooth and nail. Do we have to do this each and every day ?

So you're okay with him posting bullshit PR claims on GAF and don't want him to be held accountable for them?
 

quetz67

Banned
Penello: "Adding to that, each of our CUs is running 6% faster. It's not simply a 6% clock speed increase overall."

Ars: "What the hell does that even mean?"

Everybody would think 12CUs x 6% = 72% faster

But...wrong...every CU is 6% faster than the last one, so it is actually a whopping 89.83% speed increase.
 
Top Bottom