deleted | Bitcointalksearch.org

ares-austria

newbie

Activity: 19

Merit: 0

Quote from: abowakawa on May 24, 2023, 08:26:21 AM

Oof.

Looks like someone is suffering from a serious case of "Notice me senpai!"

No idea what that means, but sounds like some hentai stuff so not into that.

But yes, seems like Sumitomo is busy playing whack-a-mole with some simple questions.

Question 1: YalSAT beats the infinitely superior Dynex SAT Neuromorphic solver
Whack 1: Add more benchmark input that work better with Dynex SAT and "prove" it is superior.

Question 2: Parallel solver PalSAT beats Dynex SAT in their new benchmark suite
Whack 2: Delete the issue and add own optimised three thread results. Victory!

Question 3: Question why choose 3 threads to make Dynex look good and not 8 threads?
Whack 3: Modify the ticket out of context and not comment anything.

Question 4: Why delete ticket? Why not explain reasoning for choosing 3 threads for PalSAT vs. Dynex SAT?
Whack 4: Do nothing, ignore.

I mean all they have to do in this case is explain why they chose this 3 thread optimal value for their Dynex SAT solver (that was absolutely unbeatable just a week earlier). Why not include the 8 thread results to show?

Anyone with half a brain should see that there's something fishy going on here. Still looking for the big Dynex big tech reveal proving our doubts unjustified.

abowakawa

newbie

Activity: 8

Merit: 0

Oof.

Looks like someone is suffering from a serious case of "Notice me senpai!"

For everyone else who may be interested, here's the actual interview with Sumitomo:

https://www.youtube.com/watch?v=fFjxKmRsmBA

ares-austria

newbie

Activity: 19

Merit: 0

Quote from: ares-austria on May 22, 2023, 11:00:23 AM

In a quick comparison we can see that the "Dynex Neuromorphic Chip" SAT solver is actually dog slow compared to a single threaded YalSAT instance running on low-clocked server hardware: Dynex SAT solver vs. YalSAT benchmark.

Here YalSAT yielded close to 3 x performance compared to the Dynex sample. YMMV, so if you're into this kind of stuff - just do your own runs with the software:

DynexSolve (Dynex Neuromorphic Chip)
https://github.com/dynexcoin/Dynex-Neuromorphic-Chip

YalSAT
https://github.com/arminbiere/yalsat

Looks like the Dynex team has been noticing our criticism and have updated their repository with benchmarks that look to be in their favour. Unfortunately it is easy to see that they are comparing different things. The YalSAT is a single threaded Sequential Standalone Version, where as the Dynex SAT solver uses 8 threads as shown in their source code. This results in, once again, amazing results with their expanded dataset. Yes, because YalSAT is essentially outnumbered 8 to 1.

However, there is a PalSAT (Parallel Simple Portfolio Version) version included in the code repository for PalSAT. With this one the results are once again much in favor of the Local Search solver over the Dynex Solver, with PalSAT clocking in results for their dataset at over 2.6 x faster results on average. We have opened a ticket on the project for the team to clarify and include PalSAT results to their index page: Dynex SAT benchmarks for YalSAT 1.0.1 include no multithreaded results

We look forward to seeing the updated results or at least a response from the team to the public ticket.

anders9

newbie

Activity: 8

Merit: 0

Ye it's very old but works nice. Algo seems to be nice on the hardware. My uncle said he's going to try with some of his old machines too now!

Quote from: abowakawa on May 22, 2023, 08:12:10 AM

Quote from: anders9 on May 22, 2023, 06:31:58 AM

Was helpful thank you. Actually works really nice on my machine despite it being old.

Quote from: abowakawa on May 18, 2023, 08:28:20 AM

Quote from: anders9 on May 17, 2023, 02:18:39 PM

Heard this is good for GPU miners. Can I mine this on a GTX 1050 Ti?

Should work just fine. Best to check out their discord and ask the miners there. They're very helpful actually.

Heres the link: https://discord.com/invite/eQXZzMkyCm

that's great! glad to hear. it's an industrious community at the moment. so nice to see positive collaboration and ideas being thrown around and implemented. i also love that you can use your older cards with this algo.

anyway your card might be a bit old, but there is a lottery aspect to mining this one. if you hit the winning solution it can bag you an additional 1500-2500 DNX as a reward depending on the job. good luck!

https://www.youtube.com/watch?v=q3lzk3eMUkU

ares-austria

newbie

Activity: 19

Merit: 0

Quote from: ares-austria on May 22, 2023, 03:12:32 AM

Quote from: ares-austria on May 21, 2023, 10:52:18 AM

All this criticism is wiped out simply by letting industry experts vet the DynexSolve technology in public, as you agreed to.

It would seem the Dynex team is watching this space as they've organized an independent crypto youtuber to validate their claims that their SAT solver is real: https://www.youtube.com/watch?v=SkgPwYXCTeY

In a quick comparison we can see that the "Dynex Neuromorphic Chip" SAT solver is actually dog slow compared to a single threaded YalSAT instance running on low-clocked server hardware: Dynex SAT solver vs. YalSAT benchmark.

Here YalSAT yielded close to 3 x performance compared to the Dynex sample. YMMV, so if you're into this kind of stuff - just do your own runs with the software:

DynexSolve (Dynex Neuromorphic Chip)
https://github.com/dynexcoin/Dynex-Neuromorphic-Chip

YalSAT
https://github.com/arminbiere/yalsat

abowakawa

newbie

Activity: 8

Merit: 0

Quote from: anders9 on May 22, 2023, 06:31:58 AM

Was helpful thank you. Actually works really nice on my machine despite it being old.

Quote from: abowakawa on May 18, 2023, 08:28:20 AM

Quote from: anders9 on May 17, 2023, 02:18:39 PM

Heard this is good for GPU miners. Can I mine this on a GTX 1050 Ti?

Should work just fine. Best to check out their discord and ask the miners there. They're very helpful actually.

Heres the link: https://discord.com/invite/eQXZzMkyCm

that's great! glad to hear. it's an industrious community at the moment. so nice to see positive collaboration and ideas being thrown around and implemented. i also love that you can use your older cards with this algo.

anyway your card might be a bit old, but there is a lottery aspect to mining this one. if you hit the winning solution it can bag you an additional 1500-2500 DNX as a reward depending on the job. good luck!

https://www.youtube.com/watch?v=q3lzk3eMUkU

anders9

newbie

Activity: 8

Merit: 0

Was helpful thank you. Actually works really nice on my machine despite it being old.

Quote from: abowakawa on May 18, 2023, 08:28:20 AM

Quote from: anders9 on May 17, 2023, 02:18:39 PM

Heard this is good for GPU miners. Can I mine this on a GTX 1050 Ti?

Should work just fine. Best to check out their discord and ask the miners there. They're very helpful actually.

Heres the link: https://discord.com/invite/eQXZzMkyCm

ares-austria

newbie

Activity: 19

Merit: 0

Quote from: ares-austria on May 21, 2023, 10:52:18 AM

All this criticism is wiped out simply by letting industry experts vet the DynexSolve technology in public, as you agreed to.

It would seem the Dynex team is watching this space as they've organized an independent crypto youtuber to validate their claims that their SAT solver is real: https://www.youtube.com/watch?v=SkgPwYXCTeY

On the surface this seems like an incredible feat, solving something in minutes than in days (or never). However, their demo solver seems to be an implementation of a parallel local search SAT solver. They have been well researched and operate on a two common techniques; Parallel local search algorithms and the portfolio approach. A few relevant links on the topic:

The last link is a generic SAT implementation that can solve the "transformed_barthel_n_100000_r_8.000_p0_0.080_instance_001.cnf" example problem in 14.32 seconds on a low spec computer, not infinity. With some tuning it should be close to the Dynex implementation.

Since we are on the topic of SAT solvers and DMM it is worth noting that on the (now defunct) Q.Solutions site Daniel had listed a repository of a "Digital Memcomputing Machine (DMM) based SAT-Solver" on his GitHub account. Described as follows:

Quote

In “Memcomputing NP-complete problems in polynomial time using polynomial resources and
collective states“, Memcomputing is introduced as a novel non-Turing paradigm of computation
that uses interacting memory cells (memprocessors for short) to store and process information
on the same physical platform. It is claimed that it was recently proven mathematically that
memcomputing machines have the same computational power of nondeterministic Turing machines.
Therefore, they should solve NP-complete problems in polynomial time and, using the appropriate
architecture, with resources that only grow polynomially with the input size. Our implementation
reproduces the results as shown in the paper.

This repository is an implementation of the Digital Memcomputing Machine (DMM) described in
the paper with the goal to verify if the results of the paper can be reproduced as well as to
investigate if the performance extends to other SAT problem classes. Based on the boost library,
it supports multiple ODE integration schemes and has a few parameter tuning options.

Read also Physics Inspired Computation — Solving Constraint Satisfaction Problems with
Physics with details about the implementation.

Also remember to look up our notes of the potential patent infringements of Dynex vs. Memcomputing Inc.

ares-austria

newbie

Activity: 19

Merit: 0

Quote from: abowakawa on May 21, 2023, 07:35:42 AM

Oh, please. Give me a break.

You can't say that for many years now the SEC hasn't operated as a quasi protection racket for the US gov. If you can't recognize that, why are you even in crypto? Musk is best known, among many others, for publicly saying as much.

They operate with in-house judges. You can't fight them. And the case you've posted yet again... there are so many absent details. Still, on top of that, the SEC forces the accused to sign a gag order so they can't defend themselves in public.

Get a grip man.

Setting aside your passive aggression yet again, I would have agreed the Dynex PyTorch would be a great time to reveal since it will probably be very hands on.

Again, for your own credibility, feel free to tell us how you and your friends are personally driven to spend hours sniping at a small-time project at the exclusion of all others.

No need to resort to conspiracy theories and innuendo to question our credibility.

As stated there is nothing but publicly available information in our posts. This is to make sure this information is available for reference in case it proves to be useful today and/or in the years to come. There is no malice or personal vendetta, just sharing the issues that raise suspicion on the Dynex project and Daniel Mattes' involvement for us. This should have visibility as a counterweight to the Dynex marketing for people to make informed decisions based on all available information.

All this criticism is wiped out simply by letting industry experts vet the DynexSolve technology in public, as you agreed to.

abowakawa

newbie

Activity: 8

Merit: 0

Quote from: ares-austria on May 21, 2023, 01:25:38 AM

Note that this promise on Dynex' part is referring to the infrastructure, where as the Cambrian explosion of AI in the recent months has been on the application level. Also regarding the promises, look up the references to MemComputing Inc technology vs. Dynex claims from the earlier posts on this thread.

Also add in the history of Daniel Mattes, the likely identity of Dynex lead "Sumitomo": This Founder [Daniel Mattes] Just Agreed to Pay $17 Million to Settle a Fraud Charge. Now He’s Heading an A.I. Startup. That A.I. Startup is also now defunct after failing to deliver on promises. Maybe this time it is different.

Oh, please. Give me a break.

You can't say that for many years now the SEC hasn't operated as a quasi protection racket for the US gov. If you can't recognize that, why are you even in crypto? Musk is best known, among many others, for publicly saying as much.

They operate with in-house judges. You can't fight them. And the case you've posted yet again... there are so many absent details. Still, on top of that, the SEC forces the accused to sign a gag order so they can't defend themselves in public.

Get a grip man.

Setting aside your passive aggression yet again, I would have agreed the Dynex PyTorch would be a great time to reveal since it will probably be very hands on.

Again, for your own credibility, feel free to tell us how you and your friends are personally driven to spend hours sniping at a small-time project at the exclusion of all others.

ares-austria

newbie

Activity: 19

Merit: 0

Quote from: CoiNGhosT89 on May 20, 2023, 02:27:34 PM

Now with the growth in developments related to artificial intelligence technologies, this idea is not so far from reality, let's wait for more feedback on what has been happening with the project---

Note that this promise on Dynex' part is referring to the infrastructure, where as the Cambrian explosion of AI in the recent months has been on the application level. Also regarding the promises, look up the references to MemComputing Inc technology vs. Dynex claims from the earlier posts on this thread.

Also add in the history of Daniel Mattes, the likely identity of Dynex lead "Sumitomo": This Founder [Daniel Mattes] Just Agreed to Pay $17 Million to Settle a Fraud Charge. Now He’s Heading an A.I. Startup. That A.I. Startup is also now defunct after failing to deliver on promises. Maybe this time it is different.

CoiNGhosT89

newbie

Activity: 9

Merit: 0

Now with the growth in developments related to artificial intelligence technologies, this idea is not so far from reality, let's wait for more feedback on what has been happening with the project---

ares-austria

newbie

Activity: 19

Merit: 0

Quote from: abowakawa on May 18, 2023, 08:22:05 AM

If Dynex works it will be a proud achievement for crypto and computing, even Europe for that matter. Obviously you won't need to worry about doxxing at that point because they're going to haunt you. Hopefully you won't continue to begrudge them for the conclusions you've drawn today.

Yes and as told there is plenty of time for people to jump on board this corner stone of the "crypto superstructure". All that is needed is some publicly vetted proof that the technology exists and works as advertised. Vaporware is easy to develop, which is one reason why it is clear to us that Dynex is a scam, likely from Daniel Mattes.

Since they are stating to launching the the Dynex PyTorch accelerator to market soon, that would be perfect to reveal through the PyTorch community.

MMOStars

member

Activity: 289

Merit: 18

Quote from: CNMSG on May 18, 2023, 11:05:53 PM

Looking at the current price of DNX, you can tell how much money this scam team is robbing the Chinese community. Ha ha ha. None of them are innocent, they deserve it! You know you're a scammer, but believe it! Hahaha, a group of idiots. Next time, the team will quietly issue another 100 million coins to rob your wallet.Looking at the current price of DNX, you can tell how much money this scam team is robbing the Chinese community. Ha ha ha. None of them are innocent, they deserve it! You know you're a scammer, but believe it! Hahaha, a group of idiots. Next time, the team will quietly issue another 100 million coins to rob your wallet.

I got nothing related with this project, 0 coins, but I'm not seeing anything terrible with the price. People mine/people liquidating, that's about it.

https://coinmarketcap.com/currencies/dynex/

amaoback

newbie

Activity: 8

Merit: 0

CNMSB

abowakawa

newbie

Activity: 8

Merit: 0

lol anyway... moving on... Roll Eyes

CNMSG

newbie

Activity: 11

Merit: 0

A group of idiotic faith fans

CNMSG

newbie

Activity: 11

Merit: 0

Looking at the current price of DNX, you can tell how much money this scam team is robbing the Chinese community. Ha ha ha. None of them are innocent, they deserve it! You know you're a scammer, but believe it! Hahaha, a group of idiots. Next time, the team will quietly issue another 100 million coins to rob your wallet.Looking at the current price of DNX, you can tell how much money this scam team is robbing the Chinese community. Ha ha ha. None of them are innocent, they deserve it! You know you're a scammer, but believe it! Hahaha, a group of idiots. Next time, the team will quietly issue another 100 million coins to rob your wallet.

abowakawa

newbie

Activity: 8

Merit: 0

Quote from: anders9 on May 17, 2023, 02:18:39 PM

Heard this is good for GPU miners. Can I mine this on a GTX 1050 Ti?

Should work just fine. Best to check out their discord and ask the miners there. They're very helpful actually.

Heres the link: https://discord.com/invite/eQXZzMkyCm

abowakawa

newbie

Activity: 8

Merit: 0

Quote from: ares-austria on May 17, 2023, 01:59:10 PM

PS: Are you really suggesting the crypto markets are rational? Seriously.

You can't resist adding parting shots in your postscripts. That whole post it's clear you've invested a lot of ego in this and frankly it reads like a jilted ex-girlfriend. The nagging tweets don't help your case either.

In answer to your question, no I wasn't suggesting the market is rational at all. However now that you mention it, this does highlight the dichotomy within crypto investors. We aren't all moonboys, some of us spend a lot of time researching early and care about how a project will build on the crypto superstructure. A distributed crypto supercomputer has been tried many times before, but in hindsight they took obvious/orthodox approaches.

If Dynex works it will be a proud achievement for crypto and computing, even Europe for that matter. Obviously you won't need to worry about doxxing at that point because they're going to haunt you. Hopefully you won't continue to begrudge them for the conclusions you've drawn today.

Topic: deleted (Read 2214 times)