To all above, card isn't for gaming. Full stop.

totaly agree my last Quadro cost around 3000$ (on sale 🙂 ) and perform just good in gaming (mean more than GTX 1060 and less than GTX 1070)... But it isn't build for that and none gaming card can do what Quadro is intended for... Anyway i like the blue color 🙂

This is not a gaming card so the drivers are very likely terribly optimized for gaming. I think people who have these should run them through SPECviewperf , SiSoft image and cryptology, and LuxRender. These are some of the common tests for pro-directed cards. Not making excuses but AMD may not have optmzed these cards for gaming like Nvidia did on the quadros even know nobody in there right mind uses a quadro for gaming but people review them with games for some insane reason.

On average still slower. RX580 is like 2-3% faster on average than 1060 Games with poor multi-threading in DX9-11 AMD drivers are performing poor compared to pascal

Dude, don't talk without doing some research because RX480 is (+/-) over 1060 (6Gb): +7% at Sniper Elite 4 -5% at RotTR +16.5% Battlefield 1 +37.7% Hitman +11% Deus Ex MK -7% GTAV Got bored of looking through Hilbert review, but RX480 wins average 10% I'd say. RX580 is even faster than that...

p4000 $850, p5000 $2000. FE's performance is about P4000 level.

How can you say VegaFE is about Q4000 level based on a bad source?

It's throttling due to heat. There is a reason why the 375w WC edition of the card has the exact same specifications and clock speed. Also: Which is why the guy couldn't get it to run at 1600mhz stable. AMD has no problem advertising the tflops at the 1600 though - I find that amusing. http://pro.radeon.com/en-us/product/radeon-vega-frontier-edition/ https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_1060/26.html 10% is hardly "beat the crap" - and now it's 3%. https://www.techpowerup.com/reviews/Sapphire/RX_580_Nitro_Plus/30.html Which is pretty much the entire argument. That they didn't seed reviewer cards, which created this mess. Is cheaper? Uses half the power? Has application certifications? Doesn't have driver issues with professional apps? http://www.nextwarehouse.com/item/?2715581_g10e

At the time people really bashed at Polaris for what it was and although the 1060 is more expensive, people didn't care. Let's hate on AMD. Today the RX480 is about 10% faster, that's saying over a year it gained 20% performance, am I right? Plus, without this mining frenzy an RX580 would be +/-250€ wile the 1060 6Gb is +/-280€. So +/-30€ more for a card that performs worse more than 10% on average... They didn't sent reviewer cards because its for workstations... "OH BUT IT HAS GAMER MODE BLA BLA BLA"...if it hadn't people would bash AMD for it anyway...

Just ECC memory slone is a valid reason to buy a Quadro.. Plus optimized drivers.

On a Quadro maybe, but drivers where the reason number 1 I changed to the red team. And with Freesync, I'm staying!

I think a lot of these poor benchmark results have to be driver related. The P4000 is like 5 TFLOPS FP32 and the Vega Frontier is 13 TFLOPS FP32, and also supports FP16 for 25 TFLOPS in applications that can use FP16. The GPU in Vega Frontier is also going into their MI25 accelerator. Like Nvidia P100 and V100, this big Vega chip has 49bit virtual addressing and 48bit physical addressing, allowing it to have a shared memory address of 512TB and cover all the address space for a massive HSA system's GPUs. They can also talk to each other(and to AMD Epyc CPUs i believe) directly through Infinity Fabric, meaning they will be very useful in certain types of HPC applications as well as AI and machine learning. These benchmarks dont tell the full story or usefulness of these GPUs.

Hopefully it is a driver issue, but considering the target market very few may be sold until the benchmarks indicate otherwise. The MI25 accelerator group of cards will be in direct competition with the Tesla cards.

Dude, How can you say VegaFE is about Q4000 level based on a bad source?

The source is AMD labs performance benchmark numbers compared to an independent P4000 review done this year.

You're all totally ignoring fine-wine technology... Buy now to get the performance you wanted in 2-3yrs! Doh!

It is funny this see all this comments about amd being **** and Stuff and they are not taking in to account that, this is not a gaming card... lol if it performs close to and 1080 or better hahahaha is a bigggg deal

Dude, don't talk without doing some research because RX480 is (+/-) over 1060 (6Gb): +7% at Sniper Elite 4 -5% at RotTR +16.5% Battlefield 1 +37.7% Hitman +11% Deus Ex MK -7% GTAV Got bored of looking through Hilbert review, but RX480 wins average 10% I'd say. RX580 is even faster than that...

Yeah, you got bored after you selected 4/6 AMD sponsored titles including Hitman which is clearly a wonky as **** benchmark considering both cards perform identically at 1440p. The TPU test I linked you to has 20+ games and shows the 480 behind by 2-3%. That's the most comprehensive test we have.

How can you say VegaFE is about Q4000 level based on a bad source?

What is bad about the source? The VegaFE numbers come from AMD, the P4000 numbers come from multiple sites.

At the time people really bashed at Polaris for what it was and although the 1060 is more expensive, people didn't care. Let's hate on AMD. Today the RX480 is about 10% faster, that's saying over a year it gained 20% performance, am I right? Plus, without this mining frenzy an RX580 would be +/-250€ wile the 1060 6Gb is +/-280€. So +/-30€ more for a card that performs less than 10% average...

No, you're not right, for the reasons above. Also the 1060 uses way less power than the 580 and it's not the consumers fault there is a mining craze driving prices up that's mostly effecting AMD.

They didn't sent reviewer cards because its for workstations... "OH BUT IT HAS GAMER MODE BLA BLA BLA"...if it hadn't people would bash AMD for it... On a Quadro maybe, but drivers where the reason number 1 I changed to the red team. And with Freesync, I'm staying!

I've never seen anyone bash AMD for not including a gaming mode on it's WX/FirePro series. There are sites that review workstation cards. When there is this much hype behind a release, especially from your core community (gamers) you don't release a prosumer card, with a gaming mode, and expect your core audience to not take the results at face value. It's ****ty marketing, which is the argument here. I'm sure Vega will get faster with time, as I've stated several times in this thread. Here is the known issues list:

Known Issues When running ReLive Pro, the cursor is not visible. The cursor is only visible when the user has switched to gaming mode When switching between Pro Mode and Gaming Mode, Timeout Detection and Recovery (TDRs) have been observed with multiple display configurations When numerous displays are used, there is an unlikely event where the driver does not complete installation During sleep and resume, after multiple instances, a blue screen hang has been observed Specific to Linux users, infrequent system hangs have been observed with 2015 Formula One During sleep and resume, a user may experience a TDR when instant replay is enabled Intermittent stuttering has been observed with Unigine Heaven in extreme cases System hangs have been observed in stressful situation in some applications when in CrossFire mode When using 8K displays, occasional situations with no display have been observed during sleep and resume Users may encounter Bluray playback issues on PowerDVD after switching to Game Mode Infrequent hangs have been observed during wake from sleep and resume states When resume from sleep, OpenGL applications have experienced TDRs In CrossFire configurations, blue screens have been observed during stressful ACPI testing\ When rendering using Blender cycles, users may encounter infrequent iterations of poor performance With 10 bit displays, users may encounter anomalies in display results On Z170 chipsets, users may experience infrequent hangs from sleep states. Frequency may depend on system configuration. In extreme cases with stressful applications, the user may experience performance throttling in thermally stressful situations When running express install with multiple GPUs installed, the user may experience anomalies, including unexpected system reboots

That's an insane amount of critical problems for a card that's in the hands of people today. The drivers for deep learning aren't available yet and the game mode performance worst case is a mess, best case is a marketing failure. The card was clearly rushed to hit some arbitrary deadline. AMD dropped the ball on the launch, they dropped the ball on the marketing, we've yet to see if they dropped the ball on the hardware - need a few months to see that.

It is funny this see all this comments about amd being **** and Stuff and they are not taking in to account that, this is not a gaming card... lol if it performs close to and 1080 or better hahahaha is a bigggg deal

Quadro P6000 is not a gaming card, yet it outperforms a 1080Ti. Pro Duo is not a gaming card, in fact it's capable of switching between drivers the same as this - yet it performs identically to it's gaming counterpart, the Fury Nano. https://www.pcper.com/reviews/Graphics-Cards/AMD-Radeon-Pro-Duo-Review/Grand-Theft-Auto-V Not to mention that again, the vast majority of people complaining aren't even complaining about the performance of the card, but the way the release of the card was handled.

Dude, don't talk without doing some research because RX480 is (+/-) over 1060 (6Gb): +7% at Sniper Elite 4 -5% at RotTR +16.5% Battlefield 1 +37.7% Hitman +11% Deus Ex MK -7% GTAV Got bored of looking through Hilbert review, but RX480 wins average 10% I'd say. RX580 is even faster than that...

You sound triggered, take a chill pill dude. [spoiler] https://tpucdn.com/reviews/Sapphire/RX_580_Nitro_Plus/images/perfrel_2560_1440.png https://tpucdn.com/reviews/Sapphire/RX_580_Nitro_Plus/images/perfrel_2560_1440.png https://techspot-static-xjzaqowzxaoif5.stackpathdns.com/articles-info/1393/bench/Average.png

This also means the RX 480 was just 2.5% slower than the GTX 1060 at 1080p or 2fps, a negligible difference. The RX 580 boosted performance by just 5% which made it 2fps (2.5%) faster than the GTX 1060. Overall, the RX 580 was faster than the GTX 1060, albeit by an insignificant margin.

[/spoiler] I don't know where you live but here in USA 1060 is cheaper than RX480 and the RX580 has a cost increase over the RX480 already. Not to mention 1060 uses half the power I see you in every thread bashing for people saying anything negative about AMD, i think you should step outside and take a breather.

I want to see the HASH RATES 🙂 Lets see if the Miners will go for this card or not.

:roll:

The source is AMD labs performance benchmark numbers compared to an independent P4000 review done this year.

ok

You're all totally ignoring fine-wine technology... Buy now to get the performance you wanted in 2-3yrs! Doh!

From a serious standpoint, I agree with you: we shouldn't buy something based on future performance. That said, everyone is different: I tend to keep my hardware at least for 3 years, so I'll much appreciate the performance gain in 2-3 years if the price is right today.

Yeah, you got bored after you selected 4/6 AMD sponsored titles including Hitman which is clearly a wonky as **** benchmark considering both cards perform identically at 1440p. The TPU test I linked you to has 20+ games and shows the 480 behind by 2-3%. That's the most comprehensive test we have. What is bad about the source? The VegaFE numbers come from AMD, the P4000 numbers come from multiple sites. No, you're not right, for the reasons above. Also the 1060 uses way less power than the 580 and it's not the consumers fault there is a mining craze driving prices up that's mostly effecting AMD. I've never seen anyone bash AMD for not including a gaming mode on it's WX/FirePro series. There are sites that review workstation cards. When there is this much hype behind a release, especially from your core community (gamers) you don't release a prosumer card, with a gaming mode, and expect your core audience to not take the results at face value. It's ****ty marketing, which is the argument here. I'm sure Vega will get faster with time, as I've stated several times in this thread. Here is the known issues list: That's an insane amount of critical problems for a card that's in the hands of people today. The drivers for deep learning aren't available yet and the game mode performance worst case is a mess, best case is a marketing failure. The card was clearly rushed to hit some arbitrary deadline. AMD dropped the ball on the launch, they dropped the ball on the marketing, we've yet to see if they dropped the ball on the hardware - need a few months to see that. Quadro P6000 is not a gaming card, yet it outperforms a 1080Ti. Pro Duo is not a gaming card, in fact it's capable of switching between drivers the same as this - yet it performs identically to it's gaming counterpart, the Fury Nano. https://www.pcper.com/reviews/Graphics-Cards/AMD-Radeon-Pro-Duo-Review/Grand-Theft-Auto-V Not to mention that again, the vast majority of people complaining aren't even complaining about the performance of the card, but the way the release of the card was handled.

I selected some titles I play but do you want me to do a full analysis or what? The link you provided is outdated info from 1 year ago, I looked at one of Guru3D last RX580 reviews with up to date drivers. I was talking about the guy who bought one and was testing with a 550W PSU. For god sake: 550W and a tight case... Its true, the 1060 uses less power. The RX580 is more than 10% faster though. Wasn't Ryzen rushed too? Didn't it had problems at launch? Today is more polished and has more support for memory.

I don't know where you live but here in USA 1060 is cheaper than RX480 and the RX580 has a cost increase over the RX480 already.

How's Trump as president is working for you? :stewpid: PS: Here in the forum you can see my location and it's Portugal. I'd say that every Nvidia card for the past 6 or 7 years is always more expensive over AMD and performs worse.

:roll: ok From a serious standpoint, I agree with you: we shouldn't buy something based on future performance. That said, everyone is different: I tend to keep my hardware at least for 3 years, so I'll much appreciate the performance gain in 2-3 years if the price is right today. I selected some titles I play but do you want me to do a full analysis or what? The link you provided is outdated info from 1 year ago, I looked at one of Guru3D last RX580 reviews with up to date drivers. I was talking about the guy who bought one and was testing with a 550W PSU. For god sake: 550W and a tight case... Its true, the 1060 uses less power. The RX580 is more than 10% faster though. Wasn't Ryzen rushed too? Didn't it had problems at launch? Today is more polished and has more support for memory.

I provided you a link with a full analysis, Agent-A01 provided you another.. and I posted two links, one from last year (when the 480 launched) and one from this year when the 580 launched: https://www.techpowerup.com/reviews/Sapphire/RX_580_Nitro_Plus/30.htm Here is a more recent one, from May: https://www.techpowerup.com/reviews/EVGA/GTX_1080_Ti_SC2/31.html Like i said 2-3% behind according those two, mostly even at 1440p according to Agent-A01's. At this point it's pretty obvious that you're not even reading what you're responding to so I'm not wasting my time anymore.

At this point it's pretty obvious that you're not even reading what you're responding to so I'm not wasting my time anymore.

I noticed that, only reads what he wants to read.

How's Trump as president is working for you? PS: Here in the forum you can see my location and it's Portugal. I'd say that every Nvidia card for the past 6 or 7 years is always more expensive over AMD and performs worse.

See, his blood is boiling. IDK if that's supposed to be an insult but luckily I could not care less either way :infinity:

I provided you a link with a full analysis, Agent-A01 provided you another.. and I posted two links, one from last year (when the 480 launched) and one from this year when the 580 launched: https://www.techpowerup.com/reviews/Sapphire/RX_580_Nitro_Plus/30.htm Here is a more recent one, from May: https://www.techpowerup.com/reviews/EVGA/GTX_1080_Ti_SC2/31.html Like i said 2-3% behind according those two, mostly even at 1440p according to Agent-A01's. At this point it's pretty obvious that you're not even reading what you're responding to so I'm not wasting my time anymore.

Could you please check the 1st link?

IDK if that's supposed to be an insult but luckily I could not care less either way :infinity:

Of course not :banana:

According to this review: http://www.guru3d.com/articles_pages/msi_radeon_rx_580_gaming_x_review,1.html the MSI RX580 is (on average) 10% faster at 1080p compared to the 1060 6gb. It also states it consumes 42% more power. About other reviews: Everyone has different values, it's hard to chose on who to believe. Also, there are many factors that might contribute for variations. Believe what you want.

https://globenewswire.com/news-release/2017/06/27/1029535/0/en/AMD-Launches-the-World-s-Fastest-Graphics-Card-for-Machine-Learning-Development-and-Advanced-Visualization-Workloads-Radeon-Vega-Frontier-Edition-Available-Now.html https://www.digitaltrends.com/computing/amd-vega-frontier-edition-graphics/ LOL

https://globenewswire.com/news-release/2017/06/27/1029535/0/en/AMD-Launches-the-World-s-Fastest-Graphics-Card-for-Machine-Learning-Development-and-Advanced-Visualization-Workloads-Radeon-Vega-Frontier-Edition-Available-Now.html https://www.digitaltrends.com/computing/amd-vega-frontier-edition-graphics/ LOL

LOL indeed https://abload.de/img/8c4adf8d42d1fb43e7420m4sl7.png

LOL indeed https://abload.de/img/8c4adf8d42d1fb43e7420m4sl7.png

LOL what? I see 3 different cards at 3 very different pricepoints swapping places depending on which viewset you're looking at.

LOL what? I see 3 different cards at 3 very different pricepoints swapping places depending on which viewset you're looking at.

Are we looking at the same numbers? "The World’s Fastest Graphics Card" aka Vega FE gets trashed by P6000, loses handsomely against P5000, for a grand total of exactly zero wins LOL P6000 wins 8/9 except for one case in which P5000 wins. And in one or two cases, P5000 comes awfully close. And that's what CPU bound does.

I doubt any miracles will happen. Ryzen is amazing but Vega is a big nail on the coffin.

Hmmm, anyone want to buy one and send it to Hilbert to run a few days tests on it? 😉