[H]ardOCP Editor Claims Internal Problems for AMD - Polaris Underwhelming and RTG Seeking to Leave AMD For Intel

[Curufinwe_wins]

8 minutes ago, Energycore said:

GTX 1070 (aka 3/4s GP104) and 100% Polaris 10 have similar processing cores numbers and die sizes. Whether one performs better will depend on architecture nuances but that marks them as a very clear "will compete" pair.

I heard that this rumor might be debunked? That perhaps we're seeing Vega in Q1 2017 after all.  Video cites the source. Start at 11 min.

 

 

Perhaps top end Polaris 10 will hit heads with the 1070, but naturally I was speaking of full die to full die. 

 

Additionally with amd targeting the low to mid 300 dollar price point and Polaris 10 supposedly offering Fiji level performance, I have little hope of more than almost matching the 1070.

 

Also unless things turn out differently, it seems like the Polaris 10 card is going to be launched as the 480 and probably isn't a full die then yet (similar to issues with Tonga and Fiji die yields). 

 

But hey, we only have this admittedly very long nda time to worry about (with all due respect to amd, a month long nda is dramatic and not favorable contrast to the 1080 launch event). 

 

As for Vega, we really don't know I guess.  

 

Interesting times though... 

 

Typing this on my phone so sorry it if sucks to read. 

[Energycore]

1 minute ago, Curufinwe_wins said:

Perhaps top end Polaris 10 will hit heads with the 1070, but naturally I was speaking of full die to full die. 

 

Additionally with amd targeting the low to mid 300 dollar price point and Polaris 10 supposedly offering Fiji level performance, I have little hope of more than almost matching the 1070.

 

Also unless things turn out differently, it seems like the Polaris 10 card is going to be launched as the 480 and probably isn't a full die then yet (similar to issues with Tonga and Fiji die yields). 

 

But hey, we only have this admittedly very long nda time to worry about (with all due respect to amd, a month long nda is dramatic and not favorable comparison to the 1080 launch event). 

 

As for Vega, we really don't know I guess.  

 

Interesting times though... 

What date does NDA lift?

[RagnarokDel]

16 minutes ago, Curufinwe_wins said:

and we also know AMD abruptly moved Vega up by 6 months in response to the Pascal launch so sure

This is perhaps one of the 10 dumbest thing I've read about the release of a chip ever.

[Curufinwe_wins]

1 minute ago, Energycore said:

What date does NDA lift?

June 28th irrc. Very late June. 

[Curufinwe_wins]

1 minute ago, RagnarokDel said:

This is perhaps one of the 10 dumbest thing I've read about the release of a chip ever.

How so? (un) official timelines changed dramatically around pascal launch.

 

 Coincidence? 

 

Anyways @Energycore pointed out that news might have been debunked. So yea I guess we can only speculate. 

[Energycore]

1 minute ago, Curufinwe_wins said:

June 28th irrc. Very late June. 

Ouch.

 

I might even be on a trip, completely disconnected from the tech world by that point. 'Least I get to binge research everything that happened when I'm back.

[RagnarokDel]

53 minutes ago, Curufinwe_wins said:

How so? (un) official timelines changed dramatically around pascal launch.

 

 Coincidence? 

 

Anyways @Energycore pointed out that news might have been debunked. So yea I guess we can only speculate. 

Vega was always planed for early Q4 2016/q1 2017

[FirstArmada]

On 5/28/2016 at 6:20 AM, Fetzie said:

I think it is mostly conjecture.

 

I mean, take this sentence:

So he claims that Polaris runs significantly hotter than the 1080, which thermal and power throttles like crazy. Not to mention that he glosses over the fact that Polaris 10 is a competitor for the GTX 1050(ti)/1060 and Polaris 11 will likely release as competition for the 1060(ti), with a couple of balls to the wall SKUs by AIBs that might come close to the 1070. So of course it will be slower than a 1080.

 

But exactly where he is getting the thermal performance from is a mystery to me, unless he is party to NDA information which he is reporting on (he alludes to this in the forum thread attached to the article), in which case he is probably hoping that AMD sue him so that he can write another hatchet job on AMD's future ("AMD sues review website over bad reviews" would likely be the title).

 

Not to mention that his "sources" are mostly people that have left AMD (either of their own volition or fired), in which case they are hardly likely to talk about how awesome their ex-employer was.

 

Anyway, I'm not sure this is "news". It is an editorial by a writer who is anything but unbiased, and - in my opinion - belongs more in "General Discussion".

Polaris 10 is 480

Polaris 11 is 470 and lower

 

So Polaris 10 us a competitor to the 1060/ti

and Polaris 11 is the competitor to 1050/ti i

 

f they ever release a 1050 and/or 1060

[Curufinwe_wins]

16 minutes ago, RagnarokDel said:

Vega was always planed for early Q4 2016/q1 2017

It was never listed earlier than "2017" prior to the impending pascal launch, and original estimates by the industry put it at h1 2017 at the earliest.

 

Now we had rumored reports of q3 2016. Quite the big change from h1 2017.

[Trixanity]

3 hours ago, Curufinwe_wins said:

It was never listed earlier than "2017" prior to the impending pascal launch, and original estimates by the industry put it at h1 2017 at the earliest.

 

Now we had rumored reports of q3 2016. Quite the big change from h1 2017.

Rumor said October. That's Q4. Previously it was listed as Q1 2017. At best (or worst), that's a 5 month change, if we assume it has moved from March to October. It's probably more like 2-3 months. 

 

I think perhaps AMD has been taken by surprise by Nvidia's decision to launch so early but I don't think architecturally or performance-wise that they are surprised at all.

[laminutederire]

1 hour ago, Trixanity said:

 I don't think architecturally or performance-wise that they are surprised at all.

They aren't even remotely impressed by nvidia's offer if you listen to their reprentatives, which corroborate what you say

[You_are_a_cunt]

Someone hand me the tequila and lemon, 'cause there's a lot of salt that needs to get consumed

[Curufinwe_wins]

Hey if anyone is worried about a lack of neutrality, while this next article from HOCP obviously won't fix it, it was amusing...

 

http://www.hardocp.com/article/2016/05/27/founders_edition_early_adopters_paper_tax_for_20_minutes#.V02iR2grKUk

 

Quote

While I am not sure where the GeForce GTX 1080 Founders Edition is, I am sure of where it is not. And where the Founders Edition cards are not, is on etailers shelves.

That is not a totally fair description because the GTX 1080 Founders Edition was there....for something like 20 minutes this morning, then it sold out. In fact every vendor's 1080 FE cards sold out as is shown above.

 

Just the Facts Ma'am

 

OK, well if we dealt with just facts this article would probably be pretty damn boring, considering it would be to "The End" by now.

 

What we can surmise is that NVIDIA had two things in terms of inventory levels for the 1080 FE; those two things are "Jack" and "shit." To even call this a launch day is somewhat of a slap in the face to the computer hardware enthusiast. We reached out to several folks at NVIDIA for comment on actual stocking numbers that Newegg had on hand and came up with no replies at time of publishing this. Of course they would never tell us the number anyway, but we have to ask. NVIDIA's normal reply would rattle off something like, "Well Kyle, we had so many that they were falling off the truck on the way over, and since GTX 1080 is so freaking awesome there is no way we could keep up with demand. We are as surprised as you are." And of course we all know that is a fairy tale at best.

 

The real answer can be found in the slide that I took a picture of below in Austin, Texas a couple weeks ago. At the time we originally posted it, the small print went unnoticed.

 

Well that explains a lot of things! On a serious note, it explains nothing, but even though we do not have solid data on how many 1080 FE cards Newegg sold this morning, we all know it was not "a lot."

 

This brings us to the "fact" that NVIDIA did get away with charging, what has been called by many as well, an "Early Adopter's Tax for Morons and Egomaniacs," or as NVIDIA puts it, the "EAT-ME."

Rofl, that line was great.

 

Oh and from the 1070 preview article...

 

Quote

While hardware availability won't be till June 10th, today the wrapper comes off all the evaluation of performance we are able to provide you on GeForce GTX 1070 Founders Edition. We must preface with the fact that this is a "Preview" article from us today. It does not contain everything we would wish to compare to in a full review based on price. Why? Well, we literally had 4 testing days with this video card before the embargo lifted today. NVIDIA claims there was "an issue with an errant forklift" we kid you not, ouch. At any rate, we finally got our sample on Wednesday, and we've done as much as we can in the short time we had.

 

[Derangel]

19 minutes ago, Curufinwe_wins said:

Hey if anyone is worried about a lack of neutrality, while this next article from HOCP obviously won't fix it, it was amusing...

 

http://www.hardocp.com/article/2016/05/27/founders_edition_early_adopters_paper_tax_for_20_minutes#.V02iR2grKUk

 

Rofl, that line was great.

 

Oh and from the 1070 preview article...

 

 

 

One big thing to take away from both the article in the OP and the one on the Early Adopter's Tax is this: Kyle does not beat around the bush about his opinions. He's torn into multiple companies, multiple times. [H]ardOCP was one of the first companies to tear Nvidia a new asshole because of how shit the NForce 4 chipset was. They repeatedly knocked Nvidia's PCI-E bridge chip because it was a freaking oven and caused a ton of additional heat and some additional input lag on motherboards that used it. I remember back during the early Eyefinity days, Kyle would bitch about devs not supporting it properly, especially when they outright lied about it (see: Bioshock 2). And, don't forget, they were among a small handful of sites that first uncovered the massive frame pacing problems with Crossfire and kept hounding AMD to fix it.

[Curufinwe_wins]

1 hour ago, Derangel said:

 

One big thing to take away from both the article in the OP and the one on the Early Adopter's Tax is this: Kyle does not beat around the bush about his opinions. He's torn into multiple companies, multiple times. [H]ardOCP was one of the first companies to tear Nvidia a new asshole because of how shit the NForce 4 chipset was. They repeatedly knocked Nvidia's PCI-E bridge chip because it was a freaking oven and caused a ton of additional heat and some additional input lag on motherboards that used it. I remember back during the early Eyefinity days, Kyle would bitch about devs not supporting it properly, especially when they outright lied about it (see: Bioshock 2). And, don't forget, they were among a small handful of sites that first uncovered the massive frame pacing problems with Crossfire and kept hounding AMD to fix it.

 

Completely true.

 

 

 

 

 

 

 

 

 

 

 

 

[MonkeyFace]

Editorial's are not news!! 

[maulemall]

Every thing is an editorial where "Tech" news is concerned.