[EDIT]hmmm.....whats going on at AMD? (and history of 3Dfx Rampage)

[SteelCrysis]

AMD Computex event is June 1st at 10 AM local time, or May 31 at 10 PM Eastern time: http://www.anandtech.com/show/10341/amd-...16-webcast

[SteelCrysis]

R9 480 will be Polaris 10: http://videocardz.com/60232/amd-radeon-r...polaris-10
R9 480 DirectX 11 benchmarks: http://videocardz.com/60253/amd-radeon-r...benchmarks

[ocre]

Just that 3dmark?

Honestly, I do not think it would be incredibly hard to edit the reported system info. It seems very unlikely that someone has an actual Polaris card at home and then goes on to run 3dmark and sys mark while online. Much less, multiple Polaris cards...like not only a pair but then a whole variety of Polaris gpus. So they they have cut down versions and crossfire, at their house plug them in and run benchmark apps on their pc that is online...

Yeah

Not even AMD partners have Polaris chips yet. They are still in the dark. So who the hell has a box full of full gpus at their house?

It seems much more reasonable that the
Is is someone who hacked/edited the reported system info.

It is not that I think the numbers are to unrealistic. They are actually right where many people say Polaris should be. 390x ish. And that is based on and saying than Polaris will be mainstream. Of course, AMD fans would take that to the best case scenario which is 390x to below furyx. There are some who tout much higher but we will see how that goes.

I see Polaris and think it's more like tonga, perhaps at a higher clock. But, I don't think that expectations of much higher clocks are a good thing. I think clock speed has been the issue and it is why AMD has went mum. See, they were all kinds bragging and showing off Polaris way a long time ago. Before the new year, before there was any chance at all for products....we saw and rush to show off a single sample...an engineering sample fresh from the fab. They rushed it out and showed it off to the press, look we can beat Nvidia maxwell- see here. Look don't touch.

But afterwards, they went very quite. Performance per watt claims have quietly been shifted as well. From 2.5x to 2.0x when compared to their last mainstream gpus. Which I know will be a figure used against Hawaii. Polaris vs Hawaii, 2.0x the performance per watt.

I still am expecting Hawaii like performance. I have a low bar. Less than the 390x, but we will see

[SickBeast]

It all depends on the price. AMD is going to have to charge $299 or less for these cards considering how powerful the GTX 1070 is going to be.

It's a good time to be a gamer! The best news will be a price war.

[RolloTheGreat]

It all depends on the price.  AMD is going to have to charge $299 or less for these cards considering how powerful the GTX 1070 is going to be.

It's a good time to be a gamer!  The best news will be a price war.

I don't think that will help AMD.

They've been selling 290/290X for $300/less for years.

You can currently buy a 390 for $280 with a game thrown in:


http://www.newegg.com/Product/Product.as...-_-Product

Point being people who want this level of performance at that price have it already, and probably have had it for a while.

"Here's this year's 290X, lower power and cheaper for us to make!" probably won't fly with consumers.

[SickBeast]

So you think it will be cheaper then? What's your guess? $249?

If it's $199 or less AMD is going to sell a boatload of them.

[RolloTheGreat]

So you think it will be cheaper then?  What's your guess?  $249?

If it's $199 or less AMD is going to sell a boatload of them.

I hope you're right for AMDs sake, but I don't think you are.

Like I said, everyone who wants this level of AMD performance largely has bought it. If you've got $200 now, you probably had $250-$300 when the 290/290X cost that, and you probably had $250- $300 when they sold them as 390s.

You're not going to care if they're $5, been there, done that.

Contrast that with $379 for performance that used to cost $650- $1000. That will sell a ton.

Or performance 35% above that for $600-$700, that will sell a ton.

[SickBeast]

Well based on the leak that SteelCrysis showed us, it looks like Polaris could be almost as fast as a Fury X. That's pretty decent IMO. It looks to be slightly faster than a 390x. Not everyone is running high end cards like that. For me personally I would buy a card with that level of performance for $199 without hesitation. For 1080p gaming there isn't much need for much more performance than that (yet).

[ocre]

We just have to wait and see sickbeast.

To me it looks like Polaris could be Hawaii class performance, which would be pretty bad at 300 bucks because it does so little for performance price ratio. It may turn out faster but my expectations set the bar pretty low.

I think the lower the better though. Cause at least you stand the chance of being impressed. Amd has selectively chosen to market Polaris in many ways. Just take note that performance was not one of them

[RolloTheGreat]

Well based on the leak that SteelCrysis showed us, it looks like Polaris could be almost as fast as a Fury X.  That's pretty decent IMO.  It looks to be slightly faster than a 390x.  Not everyone is running high end cards like that.  For me personally I would buy a card with that level of performance for $199 without hesitation.  For 1080p gaming there isn't much need for much more performance than that (yet).

Why wouldn't you be waiting anxiously for a 1070, any flavor?

https://www.techpowerup.com/reviews/ASUS...ix/31.html

Titan X has 33% more performance at 1080p than a 390. Why would you shaft yourself to that extent over $100-$200 difference?!?

[SickBeast]

The GTX 1070 does look awesome. However Polaris at $199 with similar performance sounds even better to me. That's really wishful thinking though.

[ocre]

Yeah, am Anand tech has bashed the cap out of this card and of pascal in general. I am sure most other sites have followed suit, more obvious than ever before, there are mostly the same people posting the same crap across forum to forum.

I really can't wait till Polaris launches. How the heck can they bash and put down a card that significantly moves price vs performance when there is absolutely nothing else out to compete with it? They are saying everything from how pathetic the results are from Nvidia with a node shrink to how pascal lost so much IPC that Nvidia had to clock them high. Just whatever they can come up with. It's ridiculous. There is nothing out but pascal, which moves the bar by a very decent amount.

As for IPC recession, it's nuts. Not only do we have smaller bus and less rops, it's also common sense to know that the higher the performance-the more you become cpu bound.

Anyway, all of this is very interesting and comical. What will they say when Polaris comes out and it literally does not move the performance at all. It will barely move the perf per dollar on top of that. Performance you could have had 2 years ago for a price you could have paid 2 years ago.

I am starting to think AMD will fall significantly behind from here out. Maybe not but it does look to me like that could happen. You know these guys are not even trying to sell an AMD go anymore. They just are out sabotaging Nvidia by trying to stop people from buying their cards. They are tryin to sabotage PC gaming.

Nothing better is coming anytime soon. It is the 1070 and 1080. There is nothing else. Polaris is not competing. It will be performance from years ago. It's no where near exciting as pascal

[RolloTheGreat]

Wise words Ocre.

They have nothing to sell this time, so all they can do is hiss that NVIDIA is trying to rob you.

I have no doubts another 980Ti type card is coming- someday. It may even cost $700 if AMD releases something similar in that price range.

However; in the market of 5/27/16 the 1080 is a good deal at $700, and the 1070 is a good deal at $400.

[SteelCrysis]

Polaris NDA ends June 29: http://videocardz.com/60373/amd-polaris-...-june-29th

[SteelCrysis]

More recognition that AMD isn't competing at the high end: http://www.tomshardware.com/news/amd-pol...31879.html

[SteelCrysis]

Koduri apparently wants to split RTG from AMD: http://hardocp.com/article/2016/05/27/fr...n_futility

[SteelCrysis]

Polaris nomenclature may have changed: http://www.techpowerup.com/222894/next-g...re-changed

[SteelCrysis]

Excavator reviewed, NeoSeeker repeatedly calls it "Evacuator" on the first page: http://www.neoseeker.com/Articles/Hardwa...845/3.html

All three sets of testing data came from quad-core processors capable of running 4 threads and 64 bit data width. One difference is that the Athlon X4 845 CPU has a base operating frequency of 3.5 GHz versus the Athlon X4 880K's 4.0 GHz, and the A10-7890K APU at 4.1 GHz. The Athlon X4 845 CPU also uses Excavator microarchitecture with Carrizo cores, while the Athlon X4 880K CPU and A10-7890K APU leverage Steamroller architecture with Godavari cores.

Taking the base operating frequencies into consideration, the Athlon X4 845 CPU is at a 12.5 to 14.6 percent speed disadvantage against the other two CPUs, but the Excavator microarchitecture is touted to process 15 percent more instructions per clock cycle. Does this mean the Athlon X4 845 CPU should perform on par with the Athlon X4 880K or A10-7890K APU? Another factor is that the Carrizo core supports APU, but is turned off in the Athlon X4 845 CPU. Let’s see how these and other factors come into play with the data results.

I think when most people see, hear or read the word "fractal" they think of those kaleidoscope viewers that you hold to your eye and twist to get various images. The Apophysis test is no different in that you can manipulate the image in several aspects and produce amazing images. The file used for this test is only 12 KB in size, but usually takes over 10 minutes to complete. The three CPUs in this test all took ~12 minutes to complete, with a maximum of 30 seconds between completion times in all test runs.

The RAR test proved the oddest results in that the Athlon X4 845 is 15 seconds slower than both of the comparison CPUs. Then the Athlon X4 845 proved equal to the Athlon X4 880K while beating the A10-7890K in both ZIP tests. Handbrake is the last timed test and the results showed that processor speed does matter at times, with the overclocking tests highlighting the ~400/500 MHz speed advantage of the Athlon X4 880k and A10-7890K.

Geekbench 3, POV-Ray and Futuremark PCMark 8 all showed varying results, with no two-digit percentile differences in each. The Futuremark 3D Mark results placed the Athlon X4 845 lowest in the overall score, but best in the CPU Physics score. So does this mean the increased transistor density works to its advantage? Not if we look at the AIDA64 Queen Score that demonstrated the Athlon X4 845's low results, and its mixed results for the AIDA64 Mandel Score with the two comparison CPUs and their overclocking speeds surging ahead.

Moving to the SiSoftware Sandra tests used to benchmark the three CPUs, we find the Athlon X4 845's Dhrystone GIPS results more than 20 percent lower in the Processor Arithmetic tests, while its Whetstone GFLOPS results kept pace with the other two processors. The Athlon X4 845 also had issues with the Multi-Core Efficiency test, however its Memory Bandwidth results were over 15 percent higher than the other two CPUs. I would surmise that the first two tests relate to the lower operating speeds, whereas the second was able to better leverage the microarchitecture of the Athlon X4 845 CPU.

The last two tests of the Metro: Last Light and Batman: Arkham Origins games normally showcase nearly identical results across all processor speeds and resolutions, assuming the same video card is maintained throughout. In this case there was one difference, where the Athlon X4 845 CPU delivered ~11 percent lower frame rates at the 1920x1200 resolution for Arkham Origins. I would have to attribute this difference to the Athlon X4 845's 8-lane PCIe design.

[RolloTheGreat]

(shrugs)

$60 CPUs aren't on my radar....

[BoFox]

So.....
Anyone else notice how low the clock speed have been on all the Polaris cards we have heard of?

I have been drawn to this and know it just didn't seem right....well, at the very least it didn't look pretty to me.
Knowing that it is GCN and the consoles are updating to it, it just doesn't seem logical there will be any major gcn changes.  Polaris most likely will be very similar to gcn chips of the past, I don't expect much of an ipc improvement.

Seeing the clock speed and the core count, it seems likely polaris 10 will be pretty slow.  If they don't clock the chip running at least 1000mhz, then it probably will be about as fast as hawaii. The original 290(x).

I have been stuck on this clock speed for awhile.  Why have we seen these clocks so freaking low?

Looks like you're spot-on - there's an apparent confirmation of the finalized clock speed at a mere 1.2GHz.  (Or 1.266 GHz, to be precise:
http://videocardz.com/60824/amd-polaris-...400-series


It might be the boost clock, since the 5.83 TFLOPs count is higher than all other rumors thus far, which say 5.5 TFLOPs.  

This makes it about the same as the R9 290X, just a bit faster than the 390.  Still short of the 390X (5.9 TFLOPs, with 512-bit bus that gives 384GB/s bandwidth).  Even with 20% memory compression, 8MT/s GDDR5 running at 256-bit (yielding 256GB/s total) would still fall short, hurting the card by a couple percentage points overall - perhaps effectively undermining any performance increases from optimizations that a revised GCN architecture would give over Hawaii R9 390X. 

I really don't expect it to beat the 390X or GTX 980.  It'd probably trade blows nicely with a GTX 970, coming out ever so slightly ahead. 

Rumored price is $199, but I don't believe it.  It probably means $229 or $249 - AMD just wants to up the hype especially to hurt GTX 1070 sales.


Now that there's a rumor that Vega (AMD's next card with HBM2 memory) will be released much earlier than expected, Nvidia probably is accelerating their GP102 (gamer-oriented big Pascal) launch as well.  These might be the longest paper launches in history, if HBM2 supply isn't ready...  Nvidia might not even use HBM2 with GP102, and just stick with GDDR5X with a 384-bit bus as usual, in order to beat AMD to the punch once again, while AMD waits for HBM2.
See:  http://wccftech.com/nvidia-gtx-1080-ti-pascal-gp102/

I wouldn't be surprised if NV does GP200 shortly after GP100 is widely available - while GP200 is also available for gamers in a Geforce brand.  Probably to do away with the FP64 double precision stuff just like with Maxwell compared to Kepler.

[SteelCrysis]

http://www.extremetech.com/extreme/22972...early-2017

AMD’s Zen has been a topic of conversation for over a year now, as the chip has made its way from design schematics to functional silicon and into shipping products. Intel’s Kaby Lake refresh has been less a topic, since it’s a minor respin on existing 14nm hardware, but both companies have previously stated they would target a 2016 release date.

According to DigiTimes, however, both manufacturers will delay new product introductions supposedly due to weak market demand and upstream channel inventory management issues. While this could theoretically be true, it’s not a particularly convincing reason for either company. AMD has made it extremely clear that it will launch Zen into the PC enthusiast space first, followed by the enterprise data center. AMD currently competes in neither of these markets; the company’s FX-class processors are midrange CPUs at best (the FX-8350, at $159, is the highest-priced AMD product in the Amazon top 20).
...
The major question from enthusiasts has been when AMD will launch Zen, and the company has been a bit coy on this question. CEO Lisa Su said at Computex that Zen would begin sampling to top-tier customers this quarter and to larger OEMs in Q3. While both of these steps are important, neither of them exactly translates into a firm commitment to the retail channel.

While I have no special knowledge on this topic, I can say that both AMD and Intel typically align their channel launches with at least a handful of OEM products in order to claim that hardware is immediately available and to avoid claims of a paper launch. The other thing to remember is that very few product launches ever take place in December: Staff are typically on vacation or gearing up for CES, not prepping to kick a product out the door. A December launch is also too late to recognize any significant revenue from the new platform in the current fiscal year and it’s generally too late for people who might plan to buy a chip for Christmas.

If Lisa Su plans to sample OEMs starting August 1, AMD might still make an October / November launch window. If Q3 sampling means “September 30,” than AMD will probably hold off on retail availability. It could sample reviewers prior to a wide retail launch, or it could hold the chip until CES and debut it there for maximum impact. If I had to guess, I’d guess that AMD will either roll the chip late in the year (mid-November) for immediate retail availability, or launch it at CES 2017 for both channel customers and OEMs.

There are still rumors swirling around that Zen won’t ship until later in 2017, but these rumors appear to link back to either unclear remarks Devinder Kumar made last December, in which he implied that Zen would tape out over the next few quarters, or to confusion over the company’s data center versus consumer products roadmap. AMD has explicitly said that its data center roadmap wouldn’t launch until later in 2017, but has not stated exactly when consumer products will hit store shelves.

Zen is rumored to have targeted Haswell’s performance, but there’s still a great deal of wiggle room in that statement. The best-case scenario for AMD, of course, would be a one-to-one match against an Intel CPU, where an eight-core AMD was equivalent to an eight-core Intel. Given how badly AMD has slipped compared to its larger rival, however, this may be difficult to achieve. Even rough parity would be a huge leap forward for Sunnyvale, particularly if it manages to make up enough ground that it doesn’t have to sell its eight-core chips as competitors to Intel’s quad-cores. This is doubly true in single-threaded code, where the fastest FX-9590 can be beaten by a Core i3 in certain scenarios.

[ocre]

You know.....
Amd delaying Zen is a bad bad sign. And the reasoning, "oh uhm...we are delaying the chip because...you know....the market is slow right now...in May...so...yeah...we aren't gonna launch our revolutionary chip....it won't be out in time for back to school....or even Xmas....cause you know....the market is slow right now....see..holding off the chip, it's a very smart thing to do!"

Never mind that the PC market has taken hit after hit, year over year. Wtf!!!
The market actually was in a straight down spiral, it looked bad...hopeless. But then this sudden drop leveled off some and the situation tipped out of the spiral. It started looking better than it had in years..well, the outlook mind you.

Anyway, Q2 is always down for PC markets. But to not have the chip launch in time for Q3 or Q4? It is insane. This makes no sense what so ever. You cannot know what the back to school and Xmas quarters will look like but considering last year....they probably won't be terrible. If they truly expect it to be terrible, in a declining market...wouldn't you expect, it would be worse next year? I mean, to be so negative...to think it was so bad in 2016 that you are just gonna pit off you chip launch....2017 will follow the trend. Which leveled some but it is still not moving upward...still a downward trend...still a declining market.


It is bull shit. This is total bull shit.
Zen is not being delayed for +6months because the market was slow in may. It's bull shit.
Best case scenario- they have a manufacturer in issue, the node..re spin..something

Worst case: Zen is another dud, it's not living up to claims or expectations.

Either way.. .it's just not good news. Not at all.

[RolloTheGreat]

You nailed it Ocre. Everyone knows it, we're just in some sort of Twilight Zone reality where most of what we see online know is going to be "all is well, have faith in AMD".

Contrary to popular belief, I don't want AMD to fail and I kind of hate to see what they are going through. I was a rabid supporter for many years. (even told NVIDIA I would not buy a Core 2 to bench my 8800GTX SLi because my FX-60 should be good enough)

My only complaint with them is the evil PR team having their attack mutts follow me around getting me banned.

[SteelCrysis]

Mantle wasn't originally made by AMD, Vulkan spec is his work + some other guys: http://www.gamedev.net/topic/666419-what...try5215019

The Mantle spec is effectively written by Johan Andersson at DICE, and the Khronos Vulkan spec basically pulls Aras P at Unity, Niklas S at Epic, and a couple guys at Valve into the fold.
...
The last piece to the puzzle is that we ran out of new user-facing hardware features many years ago. Ignoring raw speed, what exactly is the user-visible or dev-visible difference between a GTX 480 and a GTX 980? A few limitations have been lifted (notably in compute) but essentially they're the same thing. MS, for all practical purposes, concluded that DX was a mature, stable technology that required only minor work and mostly disbanded the teams involved. Many of the revisions to GL have been little more than API repairs. (A GTX 480 runs full featured OpenGL 4.5, by the way.) So the reason we're seeing new APIs at all stems fundamentally from Andersson hassling the IHVs until AMD woke up, smelled competitive advantage, and started paying attention. That essentially took a three year lag time from when we got hardware to the point that compute could be directly integrated into the core of a render pipeline, which is considered normal today but was bluntly revolutionary at production scale in 2012. It's a lot of small things adding up to a sea change, with key people pushing on the right people for the right things.

[ocre]

Wow...
Thanks for that thread

SteelCrysis, that thread is a really good read. I haven't really looked into the changes going forward into these new api's, you know I only have so much time....but i found that thread way worth my tims

[RolloTheGreat]

http://www.grapheine.com/bombaytv/illust...c732e.html

[ocre]

So perhaps there will be Polaris availability....

Leaks from partners:
http://m.hexus.net/tech/news/graphics/93...itro-show/

[RolloTheGreat]

So perhaps there will be Polaris availability....

Leaks from partners:
http://m.hexus.net/tech/news/graphics/93...itro-show/

I bet you will be able to get one of those a lot easier than a 1070 or 1080. I just don't think the market is there for a $200 Radeon 290/290X.

Three years ago that level of performance was exciting.

These days Radeon 390s gather dust at $249 with a free popular game:

http://www.newegg.com/Product/Product.as...-_-Product

There have been 290s and 290Xs for $250 with 5 free games.

Point being, my guess is everyone who wanted a Radeon with that level of performance bought one over the last three years because they only cost $50 more than that.

How many people do you think have been waiting three years, watching them sell for $250 with games, and now, "It's GO TIME!" they'll leap into action because they're $50 cheaper?

The shills will get them and post benches, about how quiet and cool they are, or how well they do at Ashes or Singularity.

We'll see a proverbial crap load of posts, "I need a new video card. My budget is $199, not a penny more, and I only play Ashes of Singularity and other DX12 games. What should I buy?" from "new users" with plenty of "Welcome to the forums! You need a Polaris!" replies.

The world outside of console makers and shills will yawn.

[happy medium]

So they are releasing the 460, 470 and 480.

So basically,
The 460 4gb should perform about 370 levels using less power. $129?
The 470 4gb should perform about 380x levels using less power. $149?
The 480 8gb should perform about 390x levels using less power. $239?

And these cards will compete with the 1060 and 1050 correct?

Is this what I'm seeing?
They have to have something more!?

[ocre]

I believe you are over estimating there. My expectations: 480 will be closer to 290x reference performance. Less than the 390x. It's gonna be like tonga where performance will be up and down. Might be higher in some cases but on average it will be less that 390x. The range 290(x)~390ish.

Not consistently 390x, but less

[BoFox]

Highly likely (seems confirmed with reputable sources) that there will be 2 versions of the 480:
4GB version for $199
8GB for $229

4GB seems to be pretty good for such a weak card like this (considering that faster cards like GTX 980 and especially the Fury cards had only 4GB, and didn't do too badly).  OTOH, only $30 more for 4GB extra memory isn't too bad if one plans on keeping this card for a couple years.

I think it's a good competitive price (especially for a launch card of a new generation) - since NV would probably price their GTX 1060 anywhere from $259-299 (while it barely beats the RX 480). 
Now there's a new rumor that the 1060 will have only 192-bit bus (just like GTX 660 Ti did).  
We shall see, I guess...

[RolloTheGreat]

Highly likely (seems confirmed with reputable sources) that there will be 2 versions of the 480:
4GB version for $199
8GB for $229

4GB seems to be pretty good for such a weak card like this (considering that faster cards like GTX 980 and especially the Fury cards had only 4GB, and didn't do too badly).  OTOH, only $30 more for 4GB extra memory isn't too bad if one plans on keeping this card for a couple years.

I think it's a good competitive price (especially for a launch card of a new generation) - since NV would probably price their GTX 1060 anywhere from $259-299 (while it barely beats the RX 480). 
Now there's a new rumor that the 1060 will have only 192-bit bus (just like GTX 660 Ti did).  
We shall see, I guess...

Man, with MSI 980 Ti Golden Samples selling at newegg for $369 AMIR, why are we even talking about these hobo cards? Did we all lose our jobs over the weekend and I just haven't heard because I'm on vacation?

Seriously, you'd have to be totally poor to say,"Gosh, I better wait for this $200 meh card when I could get 37% more performance for $370".

https://www.techpowerup.com/reviews/ASUS...ng/30.html

(I know that is strix, but clocks are the same)

On sadder note- there's a flaw in my Step Up Program logic with the 980 Ti Classified I ordered for $409 (also with same clocks). Step Up doesn't include 1080FTW I want, just base clock/board models. With the 14/3 phase power board on 980Ti Classified I can probably get 1080FE level performance or close to it. (so it looks like I may be waiting for 1080Ti)

[ocre]

I don't know much about the step up program. Seems like a strange restriction though.

How much would they actually give you for your old card? Is it a lot more than what one might get if they just sold their card used?

Anyhow, I bet evga might work with you if you contact them. Just tell them you were unaware when you bought your card and how much it messes up your plan and bla bla bla

A lot of times companies will try to accommodate and some make exceptions just to please....well, most good companies do

[RolloTheGreat]

I don't know much about the step up program.  Seems like a strange restriction though.

How much would they actually give you for your old card? Is it a lot more than what one might get if they just sold their card used?

Anyhow, I bet evga might work with you if you contact them. Just tell them you were unaware when you bought your card and how much it messes up your plan and bla bla bla

A lot of times companies will try to accommodate and some make exceptions just to please....well, most good companies do

It's actually a pretty good deal, and they spell it all out very clearly on their website.

Basically you have to register your card within 14 days (free) or buy extended warranty and register in 90 days.

Then you declare desire to step up in 90 days from purchase date. Once you're in line for stock at MSRP, they notify you when available and you pay the difference in their website price and what you paid, ship them your card, they send you the new one within three days of receipt. You pay shipping both ways, so that is the only money you "lose".

If they have no stock in the 90 days, your request carries forward until they do.

They maintain a list of the cards eligible to step up to, the 1080s are currently FE, AC3 at stock speed, and a stock clock non FE/non AC3 blower card.

I suppose the AC3 one would be a legit update for me, may do that. Better than FE.

Will likely wait the 90 days in hopes 1080Ti launches.

The 980 Ti Classified is 10-15% below a stock speed 1080, KitGuru review said theirs stabilized at 1392 MHz boost clock while gaming.

I'm no hobo, but it's tough to see spending $200 to get another 15% when at sometime this year I might be able to spend another $600 and get 40% or more. The 1080 looked a lot more tempting over my reference 980 Ti than the one coming on Monday. Classified looks like a beast of a 980 Ti, I could probably OC it and get pretty close to 1080 performance for the $400.. (especially as a have a Coolermaster HAF 932 case in a 70F constant environment)

[BoFox]

Wow, $409 seems cheap for the 980 Ti Classified (faster than FTW) - it's a good clearance sale before the GP104 cards become more available.  I'm curious about how high you can get it to OC!  

There a bit less of a distance between the 1080/1070 cards and the 980 Ti, than there was between GTX 680/670 and GTX 580 (last node shrink).  So, the GP104 cards aren't as impressive as the GK104 cards were when NV stopped doing hot-clocked shaders, and this is with a full node shrink all the way from 28nm to 16nm.  I guess some of the blame goes to the GP104 chips being so small (considerably smaller than GK104 and GM204 chips), in addition to the 256-bit bus bandwidth being more stifling than ever.

Meh, NV, I'd wait for the bigger chip as well.  $699 for such a small chip..  I don't even play much games to make it worthwhile.

[RolloTheGreat]

Wow, $409 seems cheap for the 980 Ti Classified (faster than FTW) - it's a good clearance sale before the GP104 cards become more available.  I'm curious about how high you can get it to OC!  

There a bit less of a distance between the 1080/1070 cards and the 980 Ti, than there was between GTX 680/670 and GTX 580 (last node shrink).  So, the GP104 cards aren't as impressive as the GK104 cards were when NV stopped doing hot-clocked shaders, and this is with a full node shrink all the way from 28nm to 16nm.  I guess some of the blame goes to the GP104 chips being so small (considerably smaller than GK104 and GM204 chips), in addition to the 256-bit bus bandwidth being more stifling than ever.

Meh, NV, I'd wait for the bigger chip as well.  $699 for such a small chip..  I don't even play much games to make it worthwhile.

Honestly, if I had the Classified before the 1080 launched, I probably never would have sold my 980 Ti.

I promised my son I'd get him the new Ti or Titan, not the new 680.

[SteelCrysis]

Rumor that top Zen CPU will have 32 cores: http://www.extremetech.com/computing/230...rious-heat

[RolloTheGreat]

Rumor that top Zen CPU will have 32 cores: http://www.extremetech.com/computing/230...rious-heat

I can see that for servers, no clue why anyone would care at home.

[SickBeast]

Rumor that top Zen CPU will have 32 cores: http://www.extremetech.com/computing/230...rious-heat

I can see that for servers, no clue why anyone would care at home.

That chip would be a power user's dream. A lot of people do real work like 3D rendering on their PC. I still remember waiting for 3 days for my computer to render my thesis in university. At that time it was rendering about 20 high resolution images and that's how long it took on a single core overclocked Athlon Barton. I can only imagine how quickly a CPU like a 32 core Zen CPU would render 3D images.

A lot of people also do video encoding and stuff. There are uses for more cores. Not so much for gaming. As an enthusiast I would definitely love a CPU like that . I also run 32GB of RAM though.

[SteelCrysis]

This is embarrassing: http://www.techpowerup.com/forums/thread...le.223568/