Polaris 10 benchmarks...

[SelenaGomez]

I think we can all agree though. We need both companies to compete heavily so we the consumers can win. Fanboy stuff is immature and stupid.

 

[trane]

So why is Nvidia far bigger if AMD has constantly had better price/performance cards?

You have answered your own question. Few people are aware of this fact, thanks to some brilliant marketing by Nvidia and equally shoddy marketing by AMD.

 

[exar333]

I shouldn't be so happy about the ops post but I am. The op is the exact reason I have predicted Polaris to fall flat on its face. Targeting price/perf means that people actually are doing hard calculations. People rarely do that kind of math when making a purchase.

If the 1070 is 380 and p10 is 300 it won't matter if p10 is the better bang for your dollar. If the 1070 wins the general masses will buy a 1070. They are not going to read a review suite take the averages of games and realize that p10 is within x% of the 1070 but the 1070 costs y% more. I've seen like a small minority of the population.

This strategy amd has proposed works for the most analytical of people. Which is not what gamers use in purchasing decisions sorry. I mean we still pre-order as gamers....

But personally I don't care if the masses don't understand how to do math to figure out which card is the best bang for buck. Too many buy poor choice Nvidia cards, which leads to early price cuts for amd, which means even better bang for buck cards for me to use with freesync. So meh, I can't complain. I may pick up a stopgap used $300 nano if possible but probably not seeing the number in circulation.

I think AMD will be challenged this year because 'good' 1080P cards have been available for a while now (290x/290/390/390x/970/960) and P10 really just helps them with new (on the cheaper side) OEM builds. That is a great opportunity, but not one to get too excited from on the enthusiast side. The 1070 is REALLY well positioned as a card that will run pretty much everything maxed out 1440P and below. Period.

Unfortunately, the P10 sits in the middle-ground where it excels at 1080P but will not be the best for higher resolutions. This would have been a GREAT product 1-2 years ago IMHO. If this was launching side by side to Vega, it would fit in the lower side well, but I think the halo effect will matter a lot to gamers to go with 1070 if they can afford it, or just go with the 1060 when that is released. NV's strategy does make sense in that they are keeping the 960 more or less as-is and are focusing on the mid-high and higher-end SKUs to fill the performance needs of higher-res displays.

Honestly, AMD's P10 and Pascal don't really bring much at all for 1080P resolutions, other than better perf/watt (which is great) but will probably not matter as much on the desktop side. If AMD can get P10 in laptops ASAP, you could see some major design wins there however...

 

[trane]

While everyone is talking about P10, it's P11 which is going to kill it in laptops. R9 M480.

We saw GTX 950 performance at 50W 6 months back. It's surely better still. That's pretty amazing. Laptops like XPS 15 and Macbook Pro 15 are going to get a massive boost in performance - we're looking at 2x-3x.

And yes, P10 could be a good option for higher end gaming laptops too.

 

[Magee_MC]

Why be so sarcastic and why exaggerate so much? Nvidia peopel arent paying $10000 for card and why cant AMD release a fast card for cheaper?

Tesla P100 buyers are paying more than $10,000 per card.

 

[Phynaz]

Because AMD has constantly had better price/perf cards compared to Nvidia since Hawaii first came out. Hasn't done them much good as far as market share goes and people constantly want cheaper and faster or no buy. How the hell is the company supposed to grow if they have to constantly provide cheaper parts that are faster? Nvidia loves their massive profits per sale compared to AMD.

So you're upset that Nvidia is a more successful company than AMD, and you feel mocking people that buy Nvidia products accomplishes what exactly? Have you found that mocking people gets them to understand your viewpoint?

 

[RussianSensation]

Unfortunately, the P10 sits in the middle-ground where it excels at 1080P but will not be the best for higher resolutions. ...

Honestly, AMD's P10 and Pascal don't really bring much at all for 1080P resolutions, other than better perf/watt (which is great) but will probably not matter as much on the desktop side. If AMD can get P10 in laptops ASAP, you could see some major design wins there however...

You are saying having this level of performance for $249-299 at below 150W real world power usage with all the latest DP/HDMI connectors, 4K video acceleration, proper DX12 architecture isn't moving the mark for 1080p 60Hz gamers? What makes you think people upgrading to Polaris 10 are going to be R9 290/290X/390/970/980 gamers? Do you know how many people bought lower end cards as per Steam? Polaris 10 should be 50-100% faster in performance than 950/960. No offense but even if I am not excited about Polaris 11/10 and lower end cards as they are not an upgrade for me, you make it sound as if no card below GTX1070 is worth releasing in 2016-2017 because you have the $ to buy a $500 enclosure and a $600-700 graphics card?

They are just that, simple calculations to show that 1070 is not out of the reach of Polaris 10 full. But still could be really far from the unknown reality

That's now how it works at all.

#1 - We have 0 proof that 14nm FF is superior to 16nm TSMC when it comes to larger graphics chips. We have 0 proof that shows AMD was able to optimize 14nm GloFo/Samsung process more than the competitor optimized 16nm TSMC process.

#2 - A9 was the same architecture ported to 2 different nodes and TSMC didn't specifically optimize clocks for it. This comparison is incorrect. You cannot just assume that AMD's architecture is designed to clock as high as the competitor's. Therefore, you cannot at all assume that Polaris 10/Vega should clock as high as the competitor's 16nm parts by using the argument that 14nm is better or at least as good -- that's because different architectures clock differently. The optimal perf/watt for Polaris 10 could occur in the range of 1200-1350mhz, that's it.

#3 - You are assuming Polaris 10 was designed to use 161W-184W of power which is what 314mm2 competitor's cards use in games. Why do those cards have 8-pin power connectors but Polaris 10 has a single 6-pin? PCIe + 6-pin = 150W power usage maximum per spec (i.e., most likely lower in the real world).

#4 - You are assuming GCN 4.0 has superior performance per clock, perf/watt and per mm2 as the competition. That's the ONLY way the comparison of a 232mm2 6-pin chip matching a 314mm2 8-pin makes any sense. It never made sense but the entire Internet keeps repeating that somehow it does.....if Polaris can actually match x70, then it is the competitor who actually failed. Polaris 10 should not even come close to that card since it was designed to be a Mainstream product, not a High-End segment product. I don't think the competitor failed so what logical conclusion can we draw? These cards don't compete with each other like i3 doesn't compete with an i5 and i5 doesn't compete with an i7. Pretty simple.

#5 - You are assuming when AMD designed Polaris 10, it was meant to be an HD7950/7970 replacement - it is not. Notice the peak power usage of 7950/7970 and compare it to x04 parts. It aligns nicely.
https://www.techpowerup.com/reviews/AMD/HD_7950/25.html

On the contrary, Polaris 10 is a much closer HD7850/7870 Pitcairn replacement. Performance, die size, power connectors, pricing/positioning ALL prove this. It seems the entire PC gaming community has lost its mind. Not directed at you, but it's embarrassing to see this level of ignorance from so many people who supposedly follow the PC hardware landscape.

March 15, 2016:

"Meanwhile AMD has also confirmed the number of GPUs in the Vega stack and their names. We’ll be seeing a Vega 10 and a Vega 11."
http://www.anandtech.com/show/10145/amd-unveils-gpu-architecture-roadmap-after-polaris-comes-vega

It's shocking to see so many people, including professional review sites such as HardOCP, not understanding that a GPU generation has 3-4 distinct tiers of videocards.

1. Low End
2. Mainstream/Perfomrance - Polaris 10/11
3. High-end - Vega 10
4. Enthusiast - Vega 11

So instead of actually trying to genuinely learn about graphics cards for the last 1-2 decades it seems most people online haven't actually learned a thing because they are desperately trying to shove Mainstream/Performance GPU into Category 3. Then when it doesn't - because it was NEVER designed to fit there - they will claim AMD is a failure and was forced to reduce prices on Polaris 10 to compete with x04 parts.

Does it look like AMD has never told us for the last 6 months that Polaris 10/11 were always meant to be <$350 parts? Before a single x04 card showed up, we always knew Polaris 10 was never going to be a true Tahiti/Hawaii replacement. Hints about die size, 256-bit GDDR5 non-X memory, AMD's own market positioning claiming they want to bring R9 290X level of performance < $349?!

The ignorance or blatant trolling online regarding Polaris is astounding. No one is stopping anyone from buying a $380-$700 card for 1080p 60Hz gaming but Polaris 10 is a card that will give gamers good 1080p 60hz perfomrance for a much lower price. Since most gamers on Steam have lower end i3/i5 CPUs and/or older Sandy/Ivy CPUs, Polaris 10 is a perfect mainstream upgrade for budget 1080p 60Hz monitor gaming.

If you want a 1440p 60-165Hz, 3440x1440 100Hz, 1080p 120-144Hz, 4K gaming graphics card, buy a competitor or wait until Vega.

I am not trying to negatively direct my post at you but it's just mind-blowing to read so much BS online and ignorance about what 85% of PC gamers actually buy. Current and historical data shows that 85% of PC gamers don't buy $380+ videocards no matter what the rhetoric is. This could change in the future, but we have to use the data we have now. People online keep claiming that Polaris 10 is an automatic fail since it brings performance between R9 390 and Fury Air to sub-$300 price level. That's amusing to read because if I needed to build a new system once Polaris 10 launches for a 1080p 60Hz monitor, what card would actually be better than a $249-299 Polaris 10? Until the competition shows up with x60/Ti level card, there is going to be nothing better.

Finally, we have seen with many cards that comparing TFlops performance on paper is often not telling us how the card(s) perform in the real world gaming tests. For example, compare the Fury X to R9 390/290X, and you'll see what I am talking about.

So you're upset that Nvidia is a more successful company than AMD, and you feel mocking people that buy Nvidia products accomplishes what exactly? Have you found that mocking people gets them to understand your viewpoint?

No, that's not why he is upset about. He is about the utter hypocricy of the entire NV loyalist community ignoring how an after-market R9 290/290X cards were available for $250-300 since November 2014, just a $50-100 premium over the 960 2-4GB cards for most of 960's useful life, but then now the argument is used that Polaris 10 that will smash 960 by 60-75% is a worthless card because let's pay $80-150 more for a GTX1070 and enjoy the glorious $90-100 24" 1080p 60Hz PC master race gaming.

This is why objective PC gamers make fun of NV loyalists. They first ignore that a $250-280 card was ~ 60% faster than a 960 and matched 960 SLI, but now that Polaris 10 is about to offer price/performance, DX12 architecture and features that NV cannot match in the sub-$300 segment, the same people that ignored the amazing price/performance of 290 and defended 2GB VRAM on 960, are suggesting Polaris 10 is garbage. I guess releasing GTX1060/1060Ti is pointless too? No wait, it's an NV card so it'll be the greatest thing since sliced bread.

This forum would be a far better place if all the G-Sync monitor locked NV loyalists would just stay the hell out of all AMD threads and stayed there in the NV sub-forum. Most of them don't have any intentions of buying any AMD cards ever. In fact, they should celebrate AMD failing against NV as some of them claim so that they can show off their $999 mid-range x80 cards in their full glory as a status symbol.

For some reason these people cannot comprehend that a $380 1070 = 1440p 60Hz card and a cheaper Polaris 10 is a 1080p 60Hz card. All those millions of GTX750/750Ti/950/960 users didn't seem to care at all that R9 270/270X/280/280X/290 destroyed them on price/performance and performance. Now, nope, mainstream/performance PC gamers should just man up and go all the way up to a $380 1070. I have 0 problem with people actually recommending someone spend more for a faster card like the 1070 -- the problem is these same individuals never recommended such strategies during all the generations where AMD cards actually did offer exactly that (R9 270 vs. 750, R9 270X vs. 750Ti, R9 280X/380X vs. 960 2GB, R9 290 vs. 960 4GB). The hypocrisy, it's 100% real but once 1060/1060Ti launches, oh boy, the best budget gaming 1080p 60Hz card. Right, right?

 

[exar333]

That's now how it works at all.

#1 - We have 0 proof that 14nm FF is superior to 16nm TSMC when it comes to larger graphics chips. We have 0 proof that shows AMD was able to optimize 14nm GloFo/Samsung process more than the competitor optimized 16nm TSMC process.

#2 - A9 was the same architecture ported to 2 different nodes and TSMC didn't specifically optimize clocks for it. This comparison is incorrect. You cannot just assume that AMD's architecture is designed to clock as high as the competitor's. Therefore, you cannot at all assume that Polaris 10/Vega should clock as high as the competitor's 16nm parts by using the argument that 14nm is better or at least as good -- that's because different architectures clock differently. The optimal perf/watt for Polaris 10 could occur in the range of 1200-1350mhz, that's it.

#3 - You are assuming Polaris 10 was designed to use 161W-184W of power which is what 314mm2 competitor's cards use in games. Why do those cards have 8-pin power connectors but Polaris 10 has a single 6-pin? PCIe + 6-pin = 150W power usage maximum per spec (i.e., most likely lower in the real world).

#4 - You are assuming GCN 4.0 has superior performance per clock, perf/watt and per mm2 as the competition. That's the ONLY way the comparison of a 232mm2 6-pin chip matching a 314mm2 8-pin makes any sense. It never made sense but the entire Internet keeps repeating that somehow it does.....if Polaris can actually match x70, the NV failed hard. It should not even come close to that card.

#5 - You are assuming when AMD designed Polaris 10, it was meant to be an HD7950/7970 replacement - it is not. Notice the peak power usage of 7950/7970 and compare it to x04 parts. It aligns nicely.
https://www.techpowerup.com/reviews/AMD/HD_7950/25.html

On the contrary, Polaris 10 is a much closer HD7850/7870 Pitcairn replacement. Performance, die size, power connectors, pricing/positioning ALL prove this. It seems the entire PC gaming community has lost its mind. Not directed at you, but it's embarrassing to see this level of ignorance from so many people who supposedly follow the PC hardware landscape.

March 15, 2016:

"Meanwhile AMD has also confirmed the number of GPUs in the Vega stack and their names. We’ll be seeing a Vega 10 and a Vega 11."

It's shocking to see so many people, including professional review sites such as HardOCP, not understanding that a GPU generation has 3-4 distinct tiers of videocards.

1. Low End
2. Mainstream/Perfomrance - Polaris 10/11
3. High-end - Vega 10
4. Enthusiast - Vega 11

So instead of actually trying to genuinely learn about graphics cards for the last 1-2 decades it seems most people online haven't actually learned a thing because they are desperately trying to shove Mainstream/Performance GPU into Category 3. Then when it doesn't - because it was NEVER designed to fit there - they will claim AMD is a failure and was forced to reduce prices on Polaris 10 to compete with x04 parts.

Does it look like AMD has never told us for the last 6 months that Polaris 10/11 were always meant to be <$350 parts? Before a single x04 card showed up, we always knew Polaris 10 was never going to be a true Tahiti/Hawaii replacement. Hints about die size, 256-bit GDDR5 non-X memory, AMD's own market positioning claiming they want to bring R9 290X level of performance < $349?!

The ignorance or blatant trolling online regarding Polaris is astounding. No one is stopping anyone from buying a $380-$700 card for 1080p 60Hz gaming but Polaris 10 is a card that will give gamers good 1080p 60hz perfomrance for a much lower price. Since most gamers on Steam have lower end i3/i5 CPUs and/or older Sandy/Ivy CPUs, Polaris 10 is a perfect mainstream upgrade for budget 1080p 60Hz monitor gaming.

If you want a 1440p 60-165Hz, 3440x1440 100Hz, 1080p 120-144Hz, 4K gaming graphics card, buy a competitor or wait until Vega.

I am not trying to negatively direct my post at you but it's just mind-blowing to read so much BS online and ignorance about what 85% of PC gamers actually buy. Current and historical data shows that 85% of PC gamers don't buy $380+ videocards no matter what the rhetoric is. This could change in the future, but we have to use the data we have now.

Totally agree.

I think it is a lot of wishful thinking (not necessarily a bad thing) driving this, coupled with the 1070/1080 NV release. The 'Hope' was for another 4870 type chip and this just isn't being positioned as that.

 

[sm625]

That's impossible. 232mm2 chip with rumoured 1266mhz GPU clocks was never designed to beat a 314mm2 chip with 1683mhz GPU boost clocks, and also you want it to cost less?

I agree it is impossible. But I wish it did just to prove a theory. My theory is that a "Golden" Polaris, 232mm2@ 145W @ $249, even if it trades blows with a 1070 (costing $400), it still gets outsold. I am willing to bet that the 1070 still outsells the $250 Polaris by a margin of at least 4:3.That's just the state of the market. I would go even further and say that if AMD somehow miraculously managed to launch Polaris on the exact same day as the 970, the 970 still outsells it. That, unfortunately will never be proven. But we will gain some insight over the next 3 months as to just how broken this market is.

 

[Stuka87]

I keep seeing people say that Vega 11 is a higher end part than Vega 10. Yet Polaris 10 is higher end than Polaris 11. Why would AMD flip flop on what the numbers mean between these?

Or are people just wrong in saying that Vega 11 is the higher end part?

 

[Bacon1]

I keep seeing people say that Vega 11 is a higher end part than Vega 10. Yet Polaris 10 is higher end than Polaris 11. Why would AMD flip flop on what the numbers mean between these?

Or are people just wrong in saying that Vega 11 is the higher end part?

The # is the order in which they were created.

Polaris 10 first, then 11

Vega 10 first then 11.

Not related to performance, just physical creation time.

 

[RussianSensation]

Honestly, AMD's P10 and Pascal don't really bring much at all for 1080P resolutions, other than better perf/watt

Rumour:

According to WSJ article, Polaris GPUs will cost no more than 199 USD. First systems equipped with Polaris GPUs will be available end of June:

"Advanced Micro Devices Inc. is angling to lower the cost of virtual reality, targeting the field with a new line of graphics hardware priced at $199&#8212;half or less the cost of comparable products. Patrick Moorhead, an analyst with Moor Insights & Strategy briefed on AMD&#8217;s strategy, estimated that the current minimum price on cards comparable to AMD&#8217;s new models is $399. He said the $199 pricing comes as a surprise."

You care to take those words back?
http://videocardz.com/60773/amd-radeon-rx-480-to-cost-199-usd

Look, I care way more about GTX1070/1080/Vega/Big Pascal. But you gotta admit bringing R9 390 8GB level of performance to $199 with less than 150W power usage and all the latest features + DX12 + HDMI 2.0/HDR monitor support for 1080p 60Hz gamers is a BIG deal.

I am willing to bet that the 1070 still outsells the $250 Polaris by a margin of at least 4:3.

Read above.

 

[Stuka87]

Rumour:

According to WSJ article, Polaris GPUs will cost no more than 199 USD. First systems equipped with Polaris GPUs will be available end of June:

"Advanced Micro Devices Inc. is angling to lower the cost of virtual reality, targeting the field with a new line of graphics hardware priced at $199—half or less the cost of comparable products. Patrick Moorhead, an analyst with Moor Insights & Strategy briefed on AMD’s strategy, estimated that the current minimum price on cards comparable to AMD’s new models is $399. He said the $199 pricing comes as a surprise."

You care to take those words back?
http://videocardz.com/60773/amd-radeon-rx-480-to-cost-199-usd

Look, myself I care way more about GTX1070/1080/Vega/Big Pascal. But you gotta admit bringing R9 390 8GB level of performance to $199 for 1080p 60Hz gamers is a BIG deal.



Read above.

I find this very interesting:

AMD Radeon RX cards are designed for VR

It was also confirmed that Polaris GPUs will be certified for VR use by HTC and Oculus, and the graphics will be capable of what 500 USD are currently used for.

Kelt Reeves, president of Falcon Northwest Computer Systems Inc said “It’s great for getting more people into VR,”.

Sounds like they are saying a $200 RX480 will match a 980?

 

[RussianSensation]

I find this very interesting:

Sounds like they are saying a $200 RX480 will match a 980?

That's exactly it. 390X ~ 980.

2304 shaders @ 1200mhz = 5.5Tflops but Discard Accelerator, improved geometry engines will mean this little 6-pin card should match 390X/980 in next gen games.

Lower price and 2304 shaders would confirm that AMD is releasing a cut-down Polaris 10 to be able to hit those aggressive price levels. That leaves room for an RX 480X.

Now consider the context.

1080 is 54% faster than an R9 390X at 4K, 66% faster with 100% fan speed+max Power Tune and 76% faster max overclocked:
http://www.computerbase.de/2016-05/asus-geforce-gtx-1080-strix-oc-test/3/

If Polaris 2304 1200mhz = 390X using < 150W, that means a ~ 450mm2 Vega 10 part with 4096 shaders and HBM2 should be fast. There is also a rumour of Vega 10 and Vega 11 having a big difference in specs.

 

[maddie]

Why be so sarcastic and why exaggerate so much? Nvidia peopel arent paying $10000 for card and why cant AMD release a fast card for cheaper? They are the vastly smaller company and seem to be playing catch up quite often. Let be real Nvidia dominates AMD for a reason. AMD has messed up too many times and have had the same strategy for a decades now. I want AMD to pick it up. I am not a fanboy at all. I want whatever is the best bang for the buck. But AMD pisses me off

I'm confused.

How can you say this now in this thread that you started, when you wrote this yesterday in another thread you started?

Quote:

I wont be buying an AMD but i would feel weird buying a freesync monitor with good stats because its cheaper than gync ones.

http://forums.anandtech.com/showpost.php?p=38257199&postcount=1

 

[jpiniero]

Sounds like they are saying a $200 RX480 will match a 980?

I would think it would be the cutdown model, so more like 970.

 

[USER8000]

I'm confused.

How can you say this now in this thread that you started, when you wrote this yesterday in another thread you started?

Quote:

I wont be buying an AMD but i would feel weird buying a freesync monitor with good stats because its cheaper than gync ones.

http://forums.anandtech.com/showpost.php?p=38257199&postcount=1

Seems a little weird,they just opened an account this month and started making all these negative comments about the time we start getting more information about Polaris 10. It is almost like they are trying to negate the price/performance that Polaris 10 and a Freesync monitor might have,and instead they want to push people to buy a GTX1070 and a Gsync monitor instead.

It is very strange behaviour - if Polaris 10 offers very good price/performance,Nvidia will have to react to that too and this is good for all gamers and hardware enthusiasts.

Trying to negate it seems counterproductive.

 

[RussianSensation]

I wont be buying an AMD but i would feel weird buying a freesync monitor with good stats because its cheaper than gync ones.

http://forums.anandtech.com/showpost.php?p=38257199&postcount=1

Nice trolling attempt by the OP to start a thread to crap on Polaris 10 while already having made up his mind on GSync and 1070/1080

---

Polaris 10 with 1200mhz clocks:

2304 * 1.2Ghz ~ 2765 SP effective.

Add IPC gains:

5% IPC gains = ~ 2903 SP effective
10% IPC gains = ~3042 SP effective
15% IPC gains = ~3180 SP effective
20% IPC gains = ~3318 SP effective

Lowest IPC gains would already discard the 5.5Tflops figure as irrelevant when comparing it to the R9 390X/290X. And by the looks of it this isn't the full Polaris 10 die but a cut-down RX 480X part.

 

[exar333]

That's exactly it. 390X ~ 980.

2304 shaders @ 1200mhz = 5.5Tflops but Discard Accelerator, improved geometry engines will mean this little 6-pin card should match 390X/980 in next gen games.

Lower price and 2304 shaders would confirm that AMD is releasing a cut-down Polaris 10 to be able to hit those aggressive price levels. That leaves room for an RX 480X.

Now consider the context.

1080 is 54% faster than an R9 390X at 4K, 66% faster with 100% fan speed+max Power Tune and 76% faster max overclocked:
http://www.computerbase.de/2016-05/asus-geforce-gtx-1080-strix-oc-test/3/

If Polaris 2304 1200mhz = 390X using < 150W, that means a ~ 450mm2 Vega 10 part with 4096 shaders and HBM2 should be fast. There is also a rumour of Vega 10 and Vega 11 having a big difference in specs.

That is all great, but we had that performance years ago with the 290/290x (close to $200), albeit at higher power consumption. Just not that exciting...if anything, just shows how overpriced the last gen has been.

 

[IllogicalGlory]

Nice that Polaris has support for DP 1.4. Won't have to worry about an HDMI-esque debacle this time around.

 

[maddie]

Seems a little weird,they just opened an account this month and started making all these negative comments about the time we start getting more information about Polaris 10. It is almost like they are trying to negate the price/performance that Polaris 10 and a Freesync monitor might have,and instead they want to push people to buy a GTX1070 and a Gsync monitor instead.

It is very strange behaviour - if Polaris 10 offers very good price/performance,Nvidia will have to react to that too and this is good for all gamers and hardware enthusiasts.

Trying to negate it seems counterproductive.

Maybe I'm a bit paranoid but it's beyond strange. A gamer would want competition for prices in general to be as low as possible. Others however.

A lot of low post count posters are active recently.

 

[exar333]

You are saying having

Snip

Couple responses. GPUs are supposed to be fun. Hobbies to cost money and new tech is fun to play with. My purchases don't really have anything to do with P10/11. I went on the record ~6 months ago stating that my preference for my next card was going to be AMD. I wasn't pleased with the 970 debacle and I liked AMD's focus on VR and D12. The last two I cared about a lot. If performance was roughly in the same ballpark (10-15%) I would 100% have leaned toward AMD. Unfortunately, Vega isn't coming until later this year or 2017. I needed a card now...that left me with a 1070 or 1080 as my build wasn't viable until tomorrow actually.

Second, I want to point out that you are 100% correct about some of the 290x hypocrisy here. My last 2 GPU purchases (the 670 and 970) both had NV superior in the price/performance AT LAUNCH but soon after price cuts happened. I built a 7950 rig a few months after my 670 build for a family and actually did 2 builds (1-290 and 1-290x) when those got cut last summer for friends. I don't regret any of those choices...

That is why the 390/390x was such a disappointment for me. Purely looking in a vacuum, it was a re-release without some key features (like HDMI 2.0, etc.) for more $$$. Tthe 290/290x was a better buy last year and 4GB was a HUGE advantage over the 960, which I don't believe was ever a great card to buy. Release through to now, its been outclassed it's whole life.

Looking at P10 from this lens, I am disappointed. It is a solid product to market to OEMs and has great potential on mobile, but I stand by the comment that it is pretty boring for us enthusiasts here. 290/290x/970 performance has been around for almost 2 years now, and you could snag those deals for close to $250. This being $200, with some updated connections (which 4K and HDR likely won't mater much in this performance class and/price range anyway) just isn't compelling.

Full disclosure, I am a AMD stock holder (as I stated in another thread asking if posters owned stock in any/both/none of the companies (AMD/NV). I believe in their future but am waiting for Pascal to give them a go from my perspective. Until then, I would recommend their GPUs for budget builds only. They will rule the sub $200 market, but how successful has that been for them in the past 5-10 years?

 

[USER8000]

They will rule the sub $200 market, but how successful has that been for them in the past 5-10 years?

They had between 30% and 40% marketshare? On some occasions,nearly 50% marketshare??

From JPR recently:

https://jonpeddie.com/images/uploads/news/graph-pr-2rev2.png

Enthusiast graphics cards are those from $300 onwards according to JPR.

 

[exar333]

They had between 30% and 40% marketshare? On some occasions,nearly 50% marketshare??

From JPR recently:

https://jonpeddie.com/images/uploads/news/graph-pr-2rev2.png

Enthusiast graphics cards are those from $300 onwards according to JPR.

But the company actually was worse performing vs. today. That is food for thought...AMD was VERY clear last year that they were not going to be the budget brand going forward. I believe that is why the AIBs were somewhat surprised (and us) with the P10 pricing. Don't get me wrong, its great for the average Joe, but it hurts when NV can slap an extra $100 on early access for a reference design and I will go on the record and say that is probably the profit on 2-3 P10s for AMD.

Edit: I was fine paying $600+ for the x800xt 12 years ago (IIRC) and $400 for a A64 CPU. I want performance. AMD 'the bargain brand' has been disappointing to me. What was the most expensive card from AMD that I can think of rather recently? The 7970. I am pretty sure that was also the best card (along with the 5870/5850) from AMD since 2009.

 

[USER8000]

But the company actually was worse performing vs. today. That is food for thought...AMD was VERY clear last year that they were not going to be the budget brand going forward. I believe that is why the AIBs were somewhat surprised (and us) with the P10 pricing. Don't get me wrong, its great for the average Joe, but it hurts when NV can slap an extra $100 on early access for a reference design and I will go on the record and say that is probably the profit on 2-3 P10s for AMD.

Edit: I was fine paying $600+ for the x800xt 12 years ago (IIRC) and $400 for a A64 CPU. I want performance. AMD 'the bargain brand' has been disappointing to me. What was the most expensive card from AMD that I can think of rather recently? The 7970. I am pretty sure that was also the best card (along with the 5870/5850) from AMD since 2009.

AMD did fine being the bargain brand with the HD3000,HD4000,HD5000 and HD6000 series. All these series targetted price/performance.

In the last couple of years,with the HD7970,R9 290 and Fury series,they targetted the higher price ranges and high end and had more brief dips into the fastest cards territory(twice with the HD7970 and getting close to the GTX780TI and Geforce Titan with the R9 290X),and with the Fury X even managing to get close to the GTX980TI after a while. AMD produced two 400MM2+ GPUs in that period and still lost marketshare. They produced a 400MM2+ GPU in the R600 and lost marketshare. In the last four years they have been actively attacking the high end with largish GPUs and it has not worked for them.

Yet during these years they ignored the midrange and laptops more and it has seen their sales drop. Moving back to smaller and cheaper chips is the correct move for AMD.

During the period of the HD3000 to HD5000 series,they only had the fastest card once,and concentrated on price/performance and their marketshare was still between 30% to 40% and they were making reasonable profits during that time. All the chips were under 400MM2 in size.