Monday, September 4th 2017

RX Vega Achieves 43 MH/s @ 130 W in Ethereum Mining

AMD's RX Vega is more along the lines of an original computing card that was moved over to the consumer segment for gaming workloads than the other way around. Raja Koduri himself has said something along those lines (extrapolating a little more than what he can actually say), and that much can be gleaned with at least a modicum of confidence through AMD's market positioning and overall computing push. In the argument between gamers and miners, Raja Koduri didn't have all that much to say, but for AMD, a sale is a sale, and it would seem that after some tweaking, RX Vega graphics cards can achieve much increased levels of mining efficiency than their Polaris counterparts, further showing how Vega handles compute workloads much better - and more efficiently - than traditional gaming ones.
Now granted, Vega's strength in mining tasks - Ethereum in particular - stems mainly from the card's usage of HBM2 memory, as well as a wide architecture with its 4096 stream processors. By setting the core clocks to 1000 MHz, the HBM2 memory clock at 1100 MHz, and power target at -24%, Reddit user S1L3N7_D3A7H was able to leverage Vega's strengths in Ethereum's PoW (Proof of Work) algorithm, achieving 43 MH/s with just 130 W of power (104 W of these for the core alone.) For comparison, tweaked RX 580 graphics cards usually deliver around 30 MH/s with 75 W core power, which amounts to around 115 W power draw per card. So Vega is achieving 43% more hash rate with a meager 13% increase in power consumption - a worthy trade-off if miners have ever seen one. This means that Vega 64 beats RX 580 cards in single node hashrate density, meaning that miners can pack more of these cards in a single system for a denser configuration with much increased performance over a similarly specced RX 580-based mining station. This was even achieved without AMD's special-purpose Beta mining driver, which has seen reports of graphical corruption and instability - the scenario could improve for miners even more with a stable release.
Moreover, S1L3N7_D3A7H said he could probably achieve the same mining efficiency on a Vega 56, which isn't all that unbelievable - memory throughput is king in Ethereum mining, so HBm2 could still be leveraged in that graphics card. It seems that at least some of that initial Vega 64 stock went into some miner's hands, as expected. And with these news, I think we'd be forgiven for holding out to our hats in the expectation of increased Vega stock (at the original $499 for Vega 64 and $399 for Vega 56 MSRP) come October. Should the users' claims about RX Vega 56 efficiency be verified, and coeteris paribus in the mining algorithms landscape for the foreseeable future, then we can very much wait for respectable inventory until Navi enters the scene. Source: Reddit user @ S1L3N7_D3A7H
Add your own comment

102 Comments on RX Vega Achieves 43 MH/s @ 130 W in Ethereum Mining

#51
Vayra86
"Raevenlord said:
Would you say that current prices and availability are for gamers? That it's worth it to pay $800 for a Vega 64, or $600+ for a Vega 56, solely for gaming?

If you were able to purchase Vega 56 at MSRP, great. That's a pretty good gaming card, you'll be extremely satisfied - and you have objective reasons for being so.

If you purchased any Vega graphics card above MSRP, you may still feel great about your purchase, but objectively, it's almost certain that the comparable NVIDIA alternative is better in pure gaming/price/performance/power terms.

Meanwhile, if you're mining, you're actually tapping into Vega's potential and strengths, which sadly, and I would love to be wrong, isn't reflected in its gaming prowess.

That's why these aren't for gamers right now. Your mileage may vary with personal opinion, your favorite manufacturer, sure. But objectively, in a technical review, price/performance/power consumption graph like you see here at TPU, that doesn't stand.
Still, I will side with most people here who feel titles like these are on the edge of clickbait-material, or just a bit over it. I don't mind a bit of sarcasm and fun in the news posts, I really like it actually, but its a VERY fine line, and this one took it too far. I do still believe people are entitled to form their own opinion without being pushed from the onset towards a specific one.

A good contrast: in earlier articles, you used the same sort of tone of voice but ended the article with a genuine question towards the opposite. Much better that way because it opens up the debate instead of steering it.
Posted on Reply
#52
Raevenlord
News Editor
"EarthDog said:
But that also doesnt make them NOT for gaming, an inflated price.

A valid argument can be made it isnt for mining either as the ROI at its current pricing doesnt balance out with some other cards either.

See how that works... for both sides?
It does work for both sides, but sale prices impact miners less than they do gamers. For a gamer, paying $800 for a Vega 64 that brings the exact same gaming experience (within 2%) as a $520 GTX 1080 is much, much worse than for a miner, which can recoup those $ in other ways other than gaming (which the gamer almost certainly never will.)

Yes, they have lowered ROI than some other cards - but if it's still profitable for them, they'll do it, especially with these undervolts and overclocks that bring Vega's compute power to the table. even more so now; and you also have to take into account pure performance density considerations, since a single Vega is (arguably) more interesting than a pair of RX 580's, simply because you can get a single vega for 2x RX 580's, thus achieving a smaller system footprint with the same - or almost equivalent - mining power.

That's why the argument works better for one side than the other, and that's why I insist it's a much better fit for miners than gamers.

"Vayra86 said:
I don't mind a bit of sarcasm and fun in the news posts, I really like it actually, but its a VERY fine line, and this one took it too far. I do still believe people are entitled to form their own opinion without being pushed from the onset towards a specific one.

A good contrast: in earlier articles, you used the same sort of tone of voice but ended the article with a genuine question towards the opposite. Much better that way because it opens up the debate instead of steering it.
That's an excellent point, one that I can side with. While I don't feel it's clickbait, I admit that I force my own view on the matter somewhat forcefully, and immediately, with that title. And while my original interpretation of the title didn't see it that way, I understand perfectly why readers might.

As such, I will remove that excessive fat from the title, and leave this here for users to see how the change occurred.
Posted on Reply
#53
renz496
"Camm said:
People whinge about AMD abandoning gamers.

Gamers abandoned AMD a long time ago, even when its cards were faster and cheaper.

AMD's strategy will make sense in the long term - full precision shaders are unnecessary for 90% of game tasks, so the shift to 16 bit shaders will be a boon both to compute centric tasks, and gaming on those tasks.

Until then (if your one of the 30% that buys a card that's faster than a 580), either you have other shit to do than just game, or really like AMD tech no matter what.
is that so? if that's the case AMD will not gain market share when they have 6 month lead with 5870 (they even beat nvidia in market share back then). they also gain market share when they have 3 months lead on 7970. if AMD want people to buy their GPU they need the fastest single GPU crown for themselves. it was that simple.
Posted on Reply
#54
EarthDog
"Raevenlord said:
It does work for both sides, but sale prices impact miners less than they do gamers. For a gamer, paying $800 for a Vega 64 that brings the exact same gaming experience (within 2%) as a $520 GTX 1080 is much, much worse than for a miner, which can recoup those $ in other ways other than gaming (which the gamer almost certainly never will.)

Yes, they have lowered ROI than some other cards - but if it's still profitable for them, they'll do it, especially with these undervolts and overclocks that bring Vega's compute power to the table. even more so now; and you also have to take into account pure performance density considerations, since a single Vega is (arguably) more interesting than a pair of RX 580's, simply because you can get a single vega for 2x RX 580's, thus achieving a smaller system footprint with the same - or almost equivalent - mining power.

That's why the argument works better for one side than the other, and that's why I insist it's a much better fit for miners than gamers.
Better, doesnt mean good, nor does it mean it isnt a gaming card either.
Posted on Reply
#55
cucker tarlson
At least Vega is gonna have better resale value than GTX. Unless the cryptocurrency mining market plummets. Then Vega's resale value is going to go down dramatically.
Posted on Reply
#56
vega22
"Raevenlord said:
Would you say that current prices and availability are for gamers? That it's worth it to pay $800 for a Vega 64, or $600+ for a Vega 56, solely for gaming?

If you were able to purchase Vega 56 at MSRP, great. That's a pretty good gaming card, you'll be extremely satisfied - and you have objective reasons for being so.

If you purchased any Vega graphics card above MSRP, you may still feel great about your purchase, but objectively, it's almost certain that the comparable NVIDIA alternative is better in pure gaming/price/performance/power terms.

Meanwhile, if you're mining, you're actually tapping into Vega's potential and strengths, which sadly, and I would love to be wrong, isn't reflected in its gaming prowess.

That's why these aren't for gamers right now. Your mileage may vary with personal opinion, your favorite manufacturer, sure. But objectively, in a technical review, price/performance/power consumption graph like you see here at TPU, that doesn't stand.
i guess it really comes down to who you class in the "for gamers" bracket. most gamers are not spending, even, vega 56 (msrp) money as the majority of gamers buy the more entry level cards. last i was looking the 1050 was the gamers choice as it was vastly out selling all other gpu. it's us enthusiast who pay for the cards higher up the tree and are currently unable to get these cards for whatever purpose we want to use them for.


while i am not disputing the facts in the piece, i just feel the title is a large part of the reason that people are already arguing in this thread. disputes which spill out across the whole forums and do not make it the friendly place it once was :(
Posted on Reply
#57
Raevenlord
News Editor
"vega22 said:


while i am not disputing the facts in the piece, i just feel the title is a large part of the reason that people are already arguing in this thread. disputes which spill out across the whole forums and do not make it the friendly place it once was :(
I agree with that, hence my removal of the offending words from the title =)

I've made my arguments here in the comment section, and I still believe that currently Vega isn't a good option for gamers/enthusiast gamers. However, the way it was conveyed wasn't the correct one, and the one I want to have here on our site, so I prefer to take a step back instead of plowing through users' expectations.
Posted on Reply
#58
silentbogo
"Raevenlord said:
It does work for both sides, but sale prices impact miners less than they do gamers. For a gamer, paying $800 for a Vega 64 that brings the exact same gaming experience (within 2%) as a $520 GTX 1080 is much, much worse than for a miner, which can recoup those $ in other ways other than gaming (which the gamer almost certainly never will.)
The price tag impacts everyone, but I don't believe it makes it in any way a non-gaming product. I can't remember the time when "overpriced" was a barrier for PC enthusiasts. People overpay by crazy amount for everything, from constantly degenerating gaming mice to overengineered VRMs and ridiculously massive cooling solutions, from LEDs and "Limited Edition" color options to factory overclock. Even for brand names....

In case of miners - it's all about the price. Even if Vega is the best possible mining card on the market, it still does not make sense to opt for a 7-8 months ROI, when less powerful options can give you profit in 4-5 months even with current inflated prices.
Posted on Reply
#59
EarthDog
You guys need an editor who checks your articles before they publish. Too many times things needed to be changed and recanted.. I love the informed takes, a lot better than it was with other newsies in the past, but still...that title was.....
Posted on Reply
#60
Th3pwn3r
"Raevenlord said:
Would you say that current prices and availability are for gamers? That it's worth it to pay $800 for a Vega 64, or $600+ for a Vega 56, solely for gaming?

If you were able to purchase Vega 56 at MSRP, great. That's a pretty good gaming card, you'll be extremely satisfied - and you have objective reasons for being so.

If you purchased any Vega graphics card above MSRP, you may still feel great about your purchase, but objectively, it's almost certain that the comparable NVIDIA alternative is better in pure gaming/price/performance/power terms.

Meanwhile, if you're mining, you're actually tapping into Vega's potential and strengths, which sadly, and I would love to be wrong, isn't reflected in its gaming prowess.

That's why these aren't for gamers right now. Your mileage may vary with personal opinion, your favorite manufacturer, sure. But objectively, in a technical review, price/performance/power consumption graph like you see here at TPU, that doesn't stand.
I'm in total agreement. However, these days you need to be a bit more specific to prevent any counter argument since people nitpick every little thing. Vega is not for gamers in the sense that GAMERS can CURRENTLY get a lot more performance for the money out of just about any Nvidia card. Should Vega pricing get back to MSRP it's a lot more appealing, like you said 56 is actually decent, 64 still isn't good in my opinion, not unless the price comes down, power consumption producing heat being my main reason.
Posted on Reply
#61
idx
"_Flare said:
https://www.nvidia.com/content/PDF/fermi_white_papers/NVIDIA_Fermi_Compute_Architecture_Whitepaper.pdf

Efforts to exploit the GPU for non-graphical applications have been underway since 2003. By using high-level shading languages such as DirectX, OpenGL and Cg, various data parallelalgorithms have been ported to the GPU. Problems such as protein folding, stock options pricing, SQL queries, and MRI reconstruction achieved remarkable performance speedups on the GPU. These early efforts that used graphics APIs for general purpose computing were known as GPGPU programs.
While the GPGPU model demonstrated great speedups, itfaced several drawbacks. First, itrequired the programmer to possess intimate knowledge of graphics APIs and GPUarchitecture. Second, problems had to be expressed in terms of vertex coordinates, texturesand shader programs, greatly increasing program complexity. Third, basic programming features such as random reads and writes to memory were not supported, greatly restricting the programming model. Lastly, the lack of double precision support (until recently) meantsome scientific applications could not be run on the GPU.

To address these problems, NVIDIA introduced two key technologies—the G80 unified graphics and compute architecture (first introduced in GeForce 8800®, Quadro FX 5600®, andTesla C870®GPUs), and CUDA, a software and hardware architecture that enabled the GPU to be programmed with a variety of high level programming languages. Together, these two technologies represented a new way of using the GPU. Instead of programming dedicated graphics units with graphics APIs, the programmer could now write C programs with CUDA extensions and target a general purpose, massively parallel processor. We called this new way
of GPU programming “GPU Computing”—it signified broader application support, widerprogramming language support, and a clear separation from the early “GPGPU” model ofprogramming.

So GPGPU is dead, long live GPU-Computing ... (addition for the archives)
OpenCL is a GPGPU api , and since opengl 4.3 there is a shader stage called Compute Shader so any developer can throw some GP tasks on that stage of rendering .
Vulkan is also going to merge with OpenCL for GPGPU applications. (sadly all nvidia GPUs support only OpenCL 1.2 while Intel and AMD already support 2.1 and 2.2 on the way).

Khronos stated during SIGGRAPH 2017 that Vulkan is going to be not just a graphics api, Vulkan is a GPU API .
Posted on Reply
#62
TheMailMan78
Big Member
"Nephilim666 said:
No real issue with this from a consumer perspective. It would be great if AMD would market their products better so that people can be better informed.

tl;dr
AMD for mining and compute
NVIDIA for gaming
I would just be happy if AMD didn't lie about its pricing. I used to be an AMD/ATI fan. I don't take kindly to liars. How can we be better informed when even the price they tell reviewers is a blatant lie.
Posted on Reply
#63
theoneandonlymrk
"Raevenlord said:
It does work for both sides, but sale prices impact miners less than they do gamers. For a gamer, paying $800 for a Vega 64 that brings the exact same gaming experience (within 2%) as a $520 GTX 1080 is much, much worse than for a miner, which can recoup those $ in other ways other than gaming (which the gamer almost certainly never will.)

Yes, they have lowered ROI than some other cards - but if it's still profitable for them, they'll do it, especially with these undervolts and overclocks that bring Vega's compute power to the table. even more so now; and you also have to take into account pure performance density considerations, since a single Vega is (arguably) more interesting than a pair of RX 580's, simply because you can get a single vega for 2x RX 580's, thus achieving a smaller system footprint with the same - or almost equivalent - mining power.

That's why the argument works better for one side than the other, and that's why I insist it's a much better fit for miners than gamers.



That's an excellent point, one that I can side with. While I don't feel it's clickbait, I admit that I force my own view on the matter somewhat forcefully, and immediately, with that title. And while my original interpretation of the title didn't see it that way, I understand perfectly why readers might.

As such, I will remove that excessive fat from the title, and leave this here for users to see how the change occurred.
I have a waterblocked modded rx480(580modded timings++bios) and a waterblocked rx vega 64 in the same rig with enough cooling for them and a bit more and I cannot hit those values with that wattage , the memory downclocks if under powered so you have to push the slider to plus 20 to maintain that memory speed even though its undervolted and downclocked ,cards differ but id suggest hes reading the wrong value as does wccf in a rare change from their normal hyperbole.
Wccf ran their own tests that suggest that reddit users wrong as is the software he's using , hwinfo64.
I tried the latest hwinfo64 build and reported hbm memory watts drawn dropped from 100-165watts to never above 30 between builds.


And there in is the pertinent point , 90% of software does not work at all with vega regarding tuneing and monitoring software , so anyone saying anything finite about what vega can or can't do is probably missinformed including many reviewers and their overclocks , a power clamp is the only sure way I've seen of knowing vegas power draw .....

Thats a full stop comment , every software lies.
All of them are unreliable yet people are basing a lot of stuff and comments of of it, or less, word of mouth.
Posted on Reply
#64
MrGenius
Seriously!? This is 1000000000000000000000000000000000000000000000000% unsubstantiated BULLSHIT! It CANNOT be and HAS NOT been replicated or confirmed by anyone and until it is it SHOULD NOT be believed for half a second.

And YES, someone already tried to replicate and confirm the results. And came NOWHERE CLOSE!!!
http://wccftech.com/amd-rx-vega-64-pushed-43-5mhs-130w-mining-ethereum-eclipsing-polaris-efficiency-factor-2x/amp/

Proving it's total BS!!!

Let me quote myself for clarity on the matter:
"MrGenius said:
Read the article carefully. They basically proved it was impossible with the amount of watts reportedly consumed.

1 card + 248W = 43.8 MH/s
2 cards + 248W = 87 MH/s

Do the math. It doesn't add up. First, half of 248 is 124(not ~130). Second, how can 2 cards achieve pretty much exactly double the hash rate of 1 card while only consuming the exact same amount of watts as 1 card? Answer...they can't. It's BS.
https://www.techpowerup.com/forums/threads/so-long-vega.236740/
Posted on Reply
#65
qubit
Overclocked quantum bit
"EarthDog said:
Instead of taking your ball and going home with a snarky comment, how about you respond with why you feel that isnt true...its how forums should work.
I was just responding in kind, did you properly read what I said? If people think I "don't have an answer" (I do) then I don't really care either. Make a polite post to me and I'll happily discuss it. Snarky and I won't bother. Simple.

I see that a friendly and reasoned reply from Vayra, I think, has disapeared, most likely because posted during the database transition. I never had a chance to read it properly unfortunately, but I would have been happy to reply to it.
Posted on Reply
#66
B-Real
"_Flare said:
Nvidia hold back Volta because it will have tons of Compute with its 7 TPC per GPC.
The successor of the GTX1080 could have 3584 Cores with 4 GPC.
I whould bet on 5 TPC per GPC again, but with new SM (without TensorCores)
they have plenty of month for tuning now, because Vega failed so hard.
I bet GCN Navi doesn´t reach the 1080Ti either, even if Navi clocks with 2.5GHz.
What?
The Vega reached the 980Ti (1070) and even the 1080. Why are you making fake conclusions?
Posted on Reply
#67
lemkeant
Well I have a Vega 64 I just installed (had it for a couple of weeks, but waiting on my EK block). My freesync monitor kept me with AMD

I tossed it in my NCase build and the blower can barely keep up in the ITX case...FYI in case others were curious. Anyway, my rig idles around 60-70 watts. Using the same settings this person did, I'm pulling about 255 watts and the same 42-43 mh/s. Doing the math that puts the card around 180-190 watts. Definitely not 130 watts.

Attaching a super quick screenshot
Posted on Reply
#68
EarthDog
"qubit said:
I was just responding in kind, did you properly read what I said? If people think I "don't have an answer" (I do) then I don't really care either. Make a polite post to me and I'll happily discuss it. Snarky and I won't bother. Simple.

I see that a friendly and reasoned reply from Vayra, I think, has disapeared, most likely because posted during the database transition. I never had a chance to read it properly unfortunately, but I would have been happy to reply to it.
lol....

Dear qubit, can you please kindly respond to the person with whatever information you have so as to clear up the misconceptions...?

Kthx. :)
Posted on Reply
#69
Chloe Price
This toy money crazyness needs to stop. I want a card with ok price for gaming, not a mid-end card with insane price, since toy money miners buy all the cards.

Luckily I have lots of monopoly money.
Posted on Reply
#70
Camm
"renz496 said:
is that so? if that's the case AMD will not gain market share when they have 6 month lead with 5870 (they even beat nvidia in market share back then). they also gain market share when they have 3 months lead on 7970. if AMD want people to buy their GPU they need the fastest single GPU crown for themselves. it was that simple.
Nvidia still outsold AMD that half. Its the closest AMD ever came to taking a lead in marketshare, but even against the 2 series, AMD couldn't outsell Nvidia. Which is sad, because the 5xxx series were awesome.
Posted on Reply
#71
ihog6hog


This's UBIQ Coin , Not ETH Coin

UBIQ Coin use low power.
Posted on Reply
#72
qubit
Overclocked quantum bit
"EarthDog said:
lol....

Dear qubit, can you please kindly respond to the person with whatever information you have so as to clear up the misconceptions...?

Kthx. :)
That's some glorious sarcasm. :p

You didn't get my point though. ;)
Posted on Reply
#73
evernessince
"silentbogo said:
Lol. Some random guy from reddit claims 130W usage based on HWInfo screenshot and behold - it's all over the net!
I've been reading about it yesterday, and even WCFTech [!!!]... just think about it, WCFTech did a follow-up/fact checking on those claims:

In context: they are measuring RX Vega64 with 980mV undervolt at 1130/1100 MHz (vs 1000/1100 @1000mV?).

All things considered, 43.5MH/s is still an impressive result but in this context it is irrelevant. That power consumption number is total fiction of a delusional kid from reddit and until vega finally hits the shelves at promised prices - no one in their right mind is going to buy it. It is still a 6+month for a complete return on investment at MSRP, and "f^&k that" at today's fictional retail price. In terms of perf/W - a pair of undervolted GTX1060 6G's makes more sense and is abundant in stores worldwide.

So, once again, there is no reason for miners to hunt for Vega until the price drops to MSRP, the shelves are stocked, and/or AMD optimizes the crap out of it to run ~60+MH/s.
There must be allot of delusional people then, because it's been sold out since launch. FYI Vega may not be an awesome gaming card but it is still very good at professional work and mining.

":at today's fictional retail price. In terms of perf/W - a pair of undervolted GTX1060 6G's makes more sense and is abundant in stores worldwide."

Read the article, density is important to miners. In addition, GTX 1060 6GB cards are running about $300 right now.
Posted on Reply
#74
EarthDog
"qubit said:
That's some glorious sarcasm. :p

You didn't get my point though. ;)
it wasnt...and, i got it. Just looking for answers instead of playing games. ;)
Posted on Reply
#75
thesmokingman
"Recus said:
No Vega definitely isn't for gamers. AMD used Geforce cards in their Gamescom booth.


[SPOILER="pics"]
[/SPOILER]
How are they going to show A/B comparisons if they don't use a B sample? Both NV and AMD use competitor products to show A/B comparisons. OMG, this is news!
Posted on Reply
Add your own comment