yosarianilives

Okay, have a few questions about stages: amd x265 1080p ddr2 - Is the idea to have a sub each with a single core and a dual core cpu? intel wrpime 32m ddr - does s478 banned mean you're encouraging use of 4coredual and qx9xxx cpus? amd 3dm01 ddr - are we supposed to have one sub with s754 and one with socket A? catzilla 4k ddr3 amd - 2x cards or dual gpu card? Same for Nvidia 3dm06 ddr2? nvidia nightraid ddr3 - one each from 500, 600, 700 series? Superpi 32m - one each from am2 and am2+? intel ddr3 igp 3dm05 - mobile haswell? Socket adapters? intel ddr4 igp night raid - mobile tigerlake? Other iris pro? Maybe desktop only?

On similar but "more reasonable" idea, xeons and optys allowed where it makes sense?

All looks good so far, and the earlier it starts the more tc I get to enjoy lol.

Cool to see it happy on X570, renoir hated anything not b550. Now we just need 5300x so we can fullpot and see ALL 7740x golds die

So wants tight timings, interesting. Can run way high fclk than 2200 too, did you try 2500 yet?

You tease with effecient pretest, did it scale with fclk at all? Did you try more?

Overclocked the bus or just normal fluctuations?

Hopefully leeg starts announcing stages soon so we can start sourcing things that come from further away

What ic did you use for mem? Not sure if 1800 8-8-8 is "playing nice" but sounds at least not as terrible as possible lol

Yeah definitely have to have users help report them

Problem cannot be truly fixed until geek6 because the changing workload size was not designed with the understanding that cpu cache would get so big. Anyways needs to be 2gb minimum end of story.

The workload is sized depending on memory size, under 1gb the workload is 50mb and mostly fits into cache which is why it is so much faster. So now the memory latency test doesn't hit memory, it's as much a tweak as running ram cache for a storage bench thus it must be banned. 1-2gb gives workload 100mb, 2+ gives 200mb workload which will hopefully be at least a year or so before that fits in cache. We need to remove all scores with less than 2gb end of story.

Some imc can handle it cold, I'm pretty sure I had a 960t that couldn't post above 800 mhz fullpot but didn't give a shit about 4 dimms.

Doesn't look like a mobile device, if I can't run android x86 nobody can lmao

You can probably get away with it in most benches, it's what I needed for 5.5. But in Avx2, especially heavy Avx2 itsa no

What device is this?

Being stupid mostly, don't remember exactly what voltage but 1.7-1.8v. I think 1.8v was getting too much to work. Screen blacked out middle of run and since then it's good Ole 00

HP to the rescue, although chip is dead because I forgot to not be stupid when running ycruncher...

Issue is chiplet design, renoir can do this fclk too

Normal apu limits, about -110c

That will wait for zen 4 or a 5300x. It's very slow in r15 compared to 7740x, even slower if I use perf bias. Well "very slow" 300-500 mhz would be enough.

At least the oem machines are "cheap" unfortunately this hp mobo will not work with non 5000g cpus, I tested. So hard to get rid of the leftovers. Not even atx so not a simple mobo swap.

Was gonna fill the juice already but missed the fill tech at air gas by an hour. Hopefully later this week

For quad core it's revolutionary until amd releases a 5300x, unless they want to wait until 6300x to shake up quad cores

Hp oem