Reportedly less than 20,000 Vega GPUs available at launch

[Sequoia93]

http://www.tweaktown.com/news/57418/amd-radeon-rx-vega-less-20-000-available-launch/index.html

Doesn't look good for us...

[1713953582]

1713953582

[1713953582]

1713953582

[philipma1957]

http://www.tweaktown.com/news/57418/amd-radeon-rx-vega-less-20-000-available-launch/index.html

Doesn't look good for us...

May not be a big deal it looks like it is just a bit better then the Nvidia 1070

[RentGPU]

http://www.tweaktown.com/news/57418/amd-radeon-rx-vega-less-20-000-available-launch/index.html

Doesn't look good for us...

If vega turned to be that monster it will be very expensive to build even a 6 vegas rig and bad roi, or it will be good priced and not that monster just small hashrate increase and power efficient

[Sequoia93]

May not be a big deal it looks like it is just a bit better then the Nvidia 1070

How are you judging that?

It reportedly uses 250w vs a 1070's 150w, but at twice the tflops (I admittedly have no idea what that may or may not mean for mining speed)

It does look like they will be at least somewhat near in price though.

[Apneal]

Also algo's have to be optimized for HMB memory. There's huge potential to really break thru performance wise if the memory is handles correctly, but the R9 Fury/Nano has shown that no one really wants to optimize for it

[mojoxc]

Just waiting for the test unit....Then we will know what we have for mining and if the power consumption, price and hash rate to see if it even makes it appealing for mining.

[philipma1957]

May not be a big deal it looks like it is just a bit better then the Nvidia 1070

How are you judging that?

It reportedly uses 250w vs a 1070's 150w, but at twice the tflops (I admittedly have no idea what that may or may not mean for mining speed)

It does look like they will be at least somewhat near in price though.

I compared it on the bench tests in the link above.

It was close to the 1070.

Granted like you I am not sure a bench test is best to determine how a card mines.

[kemo6600]

lets say it one more time

Vega worst case will be Fury on better manufacturing process .

Either same performance with 30-50% less power  or 30-40% more performance with same power.

That is worst case.

[Marvell1]

lets say it one more time

Vega worst case will be Fury on better manufacturing process .

Either same performance with 30-50% less power  or 30-40% more performance with same power.

That is worst case.

so hashing eth at 40mhs at 150 watt?

[alucard20724]

lets say it one more time

Vega worst case will be Fury on better manufacturing process .

Either same performance with 30-50% less power  or 30-40% more performance with same power.

That is worst case.

so hashing eth at 40mhs at 150 watt?

would be nice... can currently hash eth at 31 mhs at 100 watts on 1070 with no problems

[un4given]

lets say it one more time

Vega worst case will be Fury on better manufacturing process .

Either same performance with 30-50% less power  or 30-40% more performance with same power.

That is worst case.

so hashing eth at 40mhs at 150 watt?

would be nice... can currently hash eth at 31 mhs at 100 watts on 1070 with no problems

with what settings and card brand? i have got 31 MH @ 120-130 Watt on GTX 1070 Zotac Mini. 1070 are still better @ZCASH

[kemo6600]

lets say it one more time

Vega worst case will be Fury on better manufacturing process .

Either same performance with 30-50% less power  or 30-40% more performance with same power.

That is worst case.

so hashing eth at 40mhs at 150 watt?

would be nice... can currently hash eth at 31 mhs at 100 watts on 1070 with no problems

ETH is a total different story , it needs memory not core and  HBM2 is still unknown , what we know for sure that better Fury core  will scale well for sure in Zec , pascal , DCR , core based algorithms

[alucard20724]

lets say it one more time

Vega worst case will be Fury on better manufacturing process .

Either same performance with 30-50% less power  or 30-40% more performance with same power.

That is worst case.

so hashing eth at 40mhs at 150 watt?

would be nice... can currently hash eth at 31 mhs at 100 watts on 1070 with no problems

with what settings and card brand? i have got 31 MH @ 120-130 Watt on GTX 1070 Zotac Mini. 1070 are still better @ZCASH

EVGA hybrids set to 70% 100+ core and 800+ mem... yeah, they are better for zcash, but every once in awhile eth is more profitable and they'll jump to that.    for the 1060, it would seem that eth is better most of the time.  i get 21-22 eth and 280-300 sol/s with zcash, and they prefer mining eth.  I'll probably breakout the 1060s from the 1070s this weekend to make them more balanced when mining one or the other.

[adaseb]

I have a feeling that

2017 Vega = 2014 Hawaii

When the R9 290 came out during the end of 2013, many were going crazy for the card. However it was horrible for Scrypt litecoin mining.

First it ran very hot and it had bad bios and speeds were nearly identical to the cheaper R9 280X (Tahiti).

It was maybe 10% faster but cost 50% more.

[alucard20724]

lets say it one more time

Vega worst case will be Fury on better manufacturing process .

Either same performance with 30-50% less power  or 30-40% more performance with same power.

That is worst case.

so hashing eth at 40mhs at 150 watt?

would be nice... can currently hash eth at 31 mhs at 100 watts on 1070 with no problems

ETH is a total different story , it needs memory not core and  HBM2 is still unknown , what we know for sure that better Fury core  will scale well for sure in Zec , pascal , DCR , core based algorithms

hopefully i can score a few HBM2 when they get released.   I missed the first time HBM came around, i was mining strictly with asics at the time and took a break from gpu mining.. so i have no fury cards, just before and after.

[Amph]

May not be a big deal it looks like it is just a bit better then the Nvidia 1070

How are you judging that?

It reportedly uses 250w vs a 1070's 150w, but at twice the tflops (I admittedly have no idea what that may or may not mean for mining speed)

It does look like they will be at least somewhat near in price though.

I compared it on the bench tests in the link above.

It was close to the 1070.

Granted like you I am not sure a bench test is best to determine how a card mines.

usually that is a good indicator even for mining, just compare all the bench of the previous card and see hwo they perform in mining, it's the same, but there can be few exception like with ETH

[merlinsbeard]

I really hope that AMD Vega will be comparable to the 1080. I remember the prototype running DOOM at 4k 60fps last year. I hope the leaks are for the lower end Vega card and not the higher end model.

I am considering buying the 1080 but I think I will wait since the 1080 sucks at mining.

[grrrgrrr]

HBM is quite different from GDDR5/X. This guy reports 69mh/s stock with P100 which has the same flop/s with the chip underlying Titan X Pascal/1080Ti but with HBM. Vega is HBM.

https://www.reddit.com/r/EtherMining/comments/66f2dz/nvidia_tesla_p100sxm2_x4_mining/

[spyshagg]

I have a feeling that

2017 Vega = 2014 Hawaii

When the R9 290 came out during the end of 2013, many were going crazy for the card. However it was horrible for Scrypt litecoin mining.

First it ran very hot and it had bad bios and speeds were nearly identical to the cheaper R9 280X (Tahiti).

It was maybe 10% faster but cost 50% more.

If they come any close to the 290's hawaii durability it will be great news.  I have 7 that must be mining since 2014 and are still kicking (knocks on wood)