damric

Part 3. Linpack Xtreme

Part 2. CPU pump-block mounting

Part 1. Unboxing. When I opened the parts box, oh my, what a ton of accessories for every modern socket config and a ton of wires and controllers. I'll document these in the next video. Instead of printing up an instruction video, they give you a small card with a QR code that says manual. About to see how that goes

I got it today Eventually this is going into the kids' rig. But the other day my main rig went down due to pump failure so I am so happy this arrived today so I can get back up and running until I can get a replacement dual D5.

The part Linux had to do with it is the fact that there were no Linux submissions in months, thus the weight of your single submission pulled the performance line way up. You got to remember that 99% of those chips were in some junk HP Pavilions with single channel RAM and the 35w config for 1000MHz GPU clock. It's why the performance line was pretty consistently low for Windows. When properly configured those 500 vega cores do pretty well as long as it's dual channel RAM pushed as far as the IMC will let it so that the bandwidth and infinity fabric isn't choking it. Of course they don't sell pre-built machines with 3200G and Ubuntu or whatever so yeah that Linux submission really screwed me since the Windows PPD is only like 20K on LARS. It's fair game what you did unless the rules are changed. It's still not as bad as what I accidentally did to myself back in July when I accidentally turned on the dark mode for almost a week at full overclock on my HD 7850.

K5000 is interesting to me. It's similar to what a GTX 660 or something? It's really close spec-wise to my HD7850 as far as FLOPS but it has more compute cores so I'm guessing the stock clock speed is low which makes it a really good overclocker? I just want to see a better baseline that, well...looks like a line and not a dot. Then I'd like to do battle with that with my HD7850. At one point I was considering getting the MXRT version of Pitcairn for ETF since they are pretty cheap on ebay, similar gimped clocks for reliability in medical work.

Well personally, I'm going to finally learn Linux just so my oem prebuilt can feed this 3200G at stock to LARS, which indeed is @35W configuration with a single stick of DDR4 @1866 so I can counter-espionage what the other team did to my multiplier. Yeah that whole ALL OS crap is so easily abused. In the meantime I might switch to something else for ETF. I'll let you know.

My phone camera wasn't too terrible at recording Michele's gameplay then uploading to youtube. I mean I was definitely the weakest link, as the cameraman. Here is an example of one I took about 6 months ago when she was running a Vega 56 and just got her Ryzen 5800x I have a better newer video of her actually playing and with her new RTX 3080 somewhere. I need to find it.

It's too easy to poison the database, whether it's someone running a card for 2 minutes at stock or someone running something tweaked to weaken someone else's multiplier. The painful solution is to freeze the competition for a month. Require eligible cards and backup cards to be fed to the LARS database at stock configuration for entirety of previous month (maybe 2 weeks would be long enough). Do away with the ALL OS since it's too easy to poison using linux due to far less samples. You can compare FLOPS to have an idea how cards should perform. If it seems way off, then it probably is.

Apparently all you need to do is fold on it and feed it to LARS for 2 minutes for it to be eligible (see the line graph for K5000 on LARS, it's a dot).

We need a constitutional amendment to free the slaves of these mines. FREEDOM to all GPUs of ALL BRANDS, a chicken in every pot, and a graphics card in every rig! DAMRIC2024.

On 18201 I am getting 16 minutes 53 seconds. I'll have some time tomorrow to walk you through how to tune it up.

I sure would like to see what you can get out of yours

I'm on the lookout for an InWin SR24

The network print spool updates have pretty much crippled AIRBUS, where every time they deliver a plane they need to print out half of the internet in legal forms because god forbid they use electronic copies.

Yeah it's bad enough for regular consumer users like me, but you should see the havock these updates cause at my wife's work place. They try to lock them down, block them, then one of the higher ups will let an update come through and break thousands of machines from printing or accessing email servers.

Yeah I don't know what settings the 5000+ samples were running but I imagine they were running into power limits and RAM bottlenecks, both by channel interleaving and frequency. Michele was getting like 10k ppd before with single channel.

I'll take some pictures of the BIOS settings next time I need to restart, but here is a screenshot and you can see the clocks and scary voltages:

Running 1725MHz core with 2x8GB DDR4 @3466CL16 under water with high voltages, not constrained by the weak power limits. Stock for this CPU is 1100/JEDEC, and god forbid someone were to run single channel RAM @2133MHz, the PPD would be terrible, especially if they did not unlock the APU power limit since the 35W or 65W is a shared limit for the iGPU and CPU sides. Now on the other hand, the newer VEGA 8s are OC beasts, hitting 2400MHz typically. Imagine if I get my hands on one of the new APUs with better memory controller and fabric clocks. Luckily for you, I can't quite finish a WU every day, more like 3 WUs per 5 days is what I'm estimating. When Michele ran this APU for a couple weeks I noticed her PPD was *, but she was running it at stock configurations and one stick of RAM. Then the other day after the power outage I was setting up to return with my HD 7850 on the same rig. I turned on F@H and of course it started folding on everything, so I watched, then I noticed the PPD with 2 sticks of RAM was a lot better. Then I OC'd the RAM from 2133 to 3466MT/s and it was MUCH better. Then I tuned the GPU up. 1750 seemed like the ceiling when it started to crash so I dialed it back to 1725MHz. I simply unplugged the PCIE cable from the HD 7850 after it was finished, so it's just a dead card in the loop now. The iGPU, even with massive overclocks still seems more efficient than any card I own except that GTX 750. If you are interested, I'll let you remote into the rig so you can see exactly what's goin on in there.

It's an actual person who will get pissy and try to dodge your questions about why nothing is in stock yet miners are still buying cards by the pallet direct from AIBs. Do drop them a few comments https://www.amd.com/en/direct-buy/us

I'm an Elite member and I never get a notification despite being on all of the waiting lists. But...now I figured out how to inject AMD FSR into pretty much any game so I don't think I need a GPU upgrade now

Aww hell yeah Case badges, dog tags, t-shirts, ball caps, velcro patches, pens, ect. I have a contact in Poland that 3D prints some very good case badges on Etsy for cheap. Link

Yeah I guess the last time I folded on Fermi was a decade ago with a GTX 460. I only wasted $25 on that card so I'm going to give it away if I can repair the cooler. The fans are very weak and the card runs extremely hot. If I can't repair it for next to nothing then it's going to Ethan's random bin of experiments which is the last stop before recycling

No. I had no idea. I just got into the habit of nuking the unwanted slot immediately if it started up. On one of my test benches it always tries to activate the CPU slot even when I remove that slot and save. On another note, that GTX 465 is a wash. Apparently F@H no longer works with Fermi cards due to open cl requirements. I thought they used CUDA, but the forum staff over there told me that's why it is disabled.