Mining with a tesla M2090

roz303roz303 Member Posts: 66
Building my GPU rig. Still haven't settled on a graphics card just yet. It seems the "standard" GPU of choice is an R9 280X, or similar. It is stated Radeon GPUs get better perfomance overall compared to NVidia based devices. However, in terms of a rough comparison, I did some googling to see how they compare


512 CUDA cores vs 2048 Stream processors (I realize they don't directly compare)
1.3 GHZ vs 1 GHZ clock speed
6 GB GDDR5 vs 3 GB GDDR5 memory
177 gb/s vs 288 gb/s memory bandwith
1.85 ghz vs 6 ghz memory clock speed
225 Watts vs 250 watts power consumption

It's really strange to compare the two. More memory, slower and less bandwidth, but a higher clock speed?

https://en.bitcoin.it/wiki/Why_a_GPU_mines_faster_than_a_CPU#Why_are_AMD_GPUs_faster_than_Nvidia_GPUs.3F

Scrolling down to the explanation of why AMD mines faster than Nvidia is because in the SHA256 algorithm, many shift instructions are used. Since ethash makes use of Kekkak, I decided to check what it mainly uses, which is XOR. There's still two shift instructions, one of which might require many iterations to shift. AMD fires back though with a single rotate-wherever bit align instruction. Perhaps the extra .3 GHZ could make up for this?

I'm rather unsure of whether or not the M2090 is worth the hashes/$ for $145 on ebay, whereas I could spend about $20 more to get something that can output an average of 20 MH/s.

...Then again, I read someone's benchmarks of their Quadro 4200, which has 448 CUDA cores @ 1.6 GHz, and 4 GB of GDDR5 memory @ 173 gb/s, and managed to pull around 19 MH/s. Both devices are, of course, using CUDA 2.0, and are Fermi-based architectures, so it's logical to conclude that if the 4200 can use the mining software, so can the M2090

So... slightly slower GPU with more cores, twice the memory, a slightly faster memory, and less power consumption, *MIGHT* prove advantageous over an R9280X.

I could experiment to see just what it can do. Returns within 30 days for the tesla are accepted if it doesn't work out :smile:

What do you miners think? Deal or no deal?

Comments

  • Marvell9Marvell9 Member Posts: 593 ✭✭✭
    I personally have a mix of r9 290s , r9 380s and a single r9 390 and about six 280x's

    I would have liked to the go the Nvidia route due to the lower power use but their higer end cards are just to expensive.

    its all about density if you can run six or so of the nvidias one one mother board and use less power it would be better than running the power hunger R9s for a few more mh/s of hashes.

    Plus the R9s run hot and will throttle down anyways. I went th R9 route becase I already had a bunch of those and I only don't want to manage 3-4 motherboards worth of Nvidia cards for the same hashrate of AMD cards using only 2 motherboards.
  • o0ragman0oo0ragman0o Member, Moderator Posts: 1,291 mod
    roz303, I wouldn't even go there. The high end Nvidia cards are only high end because they run specific CUDA driver features that are not accessible to the commodity cards (which use the same chips). As Ethereum POW is OpenCL, there is no benefit in those high end cards over the gamer cards.

    When I was spread sheeting GPU efficiencies I ended up dropping all the Tesla's and Quatros because their performance/cost was so low
  • archetypearchetype Member Posts: 9
    running a nVidia Tesla K10 8GB GDDR5 PCI-E x16 is the best you can do. Suitable also for games.
  • GenoilGenoil 0xeb9310b185455f863f526dab3d245809f6854b4dMember Posts: 769 ✭✭✭
    edited August 2015
    You could try Tesla M2050 on an Amazon EC2 g1 instance type, but I don't think it will perform very well set against its price. As @o0ragman0o points out, these are card that typically shine in double precision floating point ops, hence their populairity in scientific and R&D applications.

    As for the algorhytm: the Keccak/SHA256-3 stages of ethash are relatively small compared to the dagger stage. And that stage has a lot of random memory access of tiny bits of data, surrounded by a lot of XORs and MULTs. So by design the hashrate shoudl scale with memory bandwidth.

    My guess is that AMD still wins over NVidia because of faster integer ops as cards with similar memory bandwidth perform quite a bit worse. By comparsion, on the HD7850 the opencl kernel was developed on, the bandwidth is 81% of theoretical maximum. On my GTX780 with native CUDA kernel, I get only 47%. This means the NVidia cards cannot fully hide memory latency because of slower integer ops. Maybe it can be improved by increasing thread occupancy, which is about 50%. Don't have any comparison on AMD.

    I would to try out ethashing on a Tesla K80 though. 100% thread occupancy is possible there because of the double register size. My request to NVidia for a test drive was denied however :)

    Here's some graphs from NSight to illustrate the bandwidth dependency:


    And this gives an idea about distribution of ALU workload. XOR fall within the Logical Ops. Shuffle has nothing to do with shifting btw. It's some CUDA specific thing that I tried to take advantage of to no avail :)




  • o0ragman0oo0ragman0o Member, Moderator Posts: 1,291 mod
    Genoil said:

    ...
    As for the algorhytm: the Keccak/SHA256-3 stages of ethash are relatively small compared to the dagger stage. And that stage has a lot of random memory access of tiny bits of data, surrounded by a lot of XORs and MULTs. So by design the hashrate shoudl scale with memory bandwidth.
    ...

    @Genoil, I was more than mildly surprised at the poor performance with the Fury's given the extrodinary bandwidth of Radeon's HBM over DDR5. I'm still clueless as to why they barely hash better than a 290x and worse than a 390x. I'm still thinking they're only running at half the performance and imagine some dedicated kernel work could prove highly profitable.
  • GenoilGenoil 0xeb9310b185455f863f526dab3d245809f6854b4dMember Posts: 769 ✭✭✭
    edited August 2015
    @o0ragman0o I'm a bit out of my comfort zone here, but it may just be the case that Fury's HBM and ethash aren't best friends due to the nature of ethash. Other than normal GPU memory reads, that usually consist of large chunks of coalesced data (textures, vertices), ethash sequentically picks tiny fractions (128 bits) from random locations in the DAG. It might therefore not be able to optimally use to ultra-wide 4096 bit bus and then suffers from the relatively low clocks.
  • roz303roz303 Member Posts: 66
    edited August 2015

    roz303, I wouldn't even go there. The high end Nvidia cards are only high end because they run specific CUDA driver features that are not accessible to the commodity cards (which use the same chips). As Ethereum POW is OpenCL, there is no benefit in those high end cards over the gamer cards.

    When I was spread sheeting GPU efficiencies I ended up dropping all the Tesla's and Quatros because their performance/cost was so low

    High end as of four years ago LMAO. The only reason I'm considering it is because they're selling hella cheap right now, and they've got a little more under the hood than a Quadro 4200, which already does a similar hashrate the 280X does, according to a benchmark on this site. Plus the tesla is less power hungry, and a little more fun to say you mine with one!

    There's Genoil's Cudaminer - is that not usable for frontier?

    Also, wouldn't you think that while the 280X's memory bandwidth is nearly twice that of the M2090, there'd be an advantage with the M2090 having twice the memory?

    I'm still most curious as to why the Quadro 4200 performs so well. Perhaps it's the extra GB of memory compared to the 280X?
  • roz303roz303 Member Posts: 66
    You know what I just realized? That benchmark I keep referring to *might* actually be a K4200 - because there's another benchmark for Genoil's cudaminer with a GTX 970, which has many more cuda cores than the M2090, and only gets 17 MHs. And, of course, the K4200 still has many more cores than the M2090... If that's the case, I'll save my money for a 7990. Ah well :wink:
  • Marvell9Marvell9 Member Posts: 593 ✭✭✭
    you cant buy a 7990 anywhere look around no one is selling them
Sign In or Register to comment.