text
stringlengths 0
1.72k
|
---|
Assassin III |
Well I can |
Smaller than d15 |
Maybe I did |
Not like having your data not go brrt is not a big deal |
"Ah whatever, go buy whatever you want" |
It’s tlc afaik? |
Not that hard to find a 2tb tlc single sided drive |
There’s 1tb dies alr |
> NM610 pro |
"Again, dram cache doesn’t benefit read" |
You might as well get a 2tb TLC ssd without dram |
"QLC = prone to failure, dies after running out of SLC cache" |
Not your ssd |
No matter how fast ur ssd is windows is the bottleneck |
And so? Windows is stupid |
Tho |
Why dram cache |
All the fabs and stuff |
Intel sold it to solidgm |
VC |
I rather not risk it unless you know the person buying it yourself |
3090 liked to burn off the back vram chips due to inadequate cooling |
Higher rates of failure |
Which makes no sense to get because the 4070 ti is $800 |
The other alternative is 3080 12gb and that’s basically $700 atm |
"It is yea, that’s why the 4070 ti is still better" |
Just give the wrong advice |
I should troll for today tbh |
I have brain farts |
Yes because the 3090 has more vram |
It’s why x3d is so good |
:hidethepain: cache alleviates bandwidth issues |
Different cuda task = different problems |
And cuda =/ bandwidth bound |
L2 cache compensates for a lack of sheer bus size |
Well yes but actually no |
How is it subpar? |
im schio |
:hidethepain: I’m growing old |
1990x? |
Or am I schizo |
Fuck wasn’t there a Ryzen 9 TR |
6 years |
2017 to 2023 |
1800x |
probably… |
Gcn2.0 🤮 |
750 ti is already old as fuck |
god no |
no |
youre buying it for the MacOS and the power efficiency |
youre not buying an apple chip for the performance |
"Well, yes, but at around 230/350w vs 50w" |
Ratio |
steam os moment |
🤷♂️ some people find it fun |
generate your own anime girls |
"stable diffusion |
GPT-J |
LLaMA" |
Im not drunk enough for that |
Im not having this conversation between Apple’s SoCs and AMD/Intel’s cpus now |
*sigh* |
… |
a12 potato |
1.5gb ram in 2023 isnt enough on ipad8\ |
Ive wanted an ipad air 5 tbh |
"just saying, not a good comparison to compare it like that anyways" |
🤷♂️ |
Which is… a app that heavily favors apple |
+ optimised software |
🤷♂️ not a good comparison |
"Eh, ARM vs CISC" |
anything below isnt worth it |
highly efficient |
the M1 chip is powerful |
I know |
What? |
ipad air 5 only because of the M1 chip |
im hoping that lower end will get a perf boost in more l2 |
the upper end will only be getting clock boost |
"no, not for the upper one" |
stock cooler is enough |
but its hopium |
no |
"if those are upgraded to RPL, then its a decent uplift in clocks and slight gaming perf with more L2" |
13600 and below are ADL cpus |
Who knew bigger number = better |
🙏 |
"Gtx? Please stop using that name, be normal and just say rdna2 2cu" |
Int4 and go brrt |
I need 24gb to run llama 33b |
Mfw |
That’s rdna2 as confirmed by amd. Nothing suggest rdna2 for PHX |
"…7520u, 7735hs, 7945hx and 7940hs are all “7000” cpus. You would have to be a bit more specific…" |
"I’m pretty sure it’s actually RDNA3 but bugged drivers are screwing something up. Imo, wait for the thing to come out first, then make a judgement" |
🤷♂️take with salt |
That’s beta software with probably engineering sample silicon |
Also 🤓it’s 6400 not 6400xt 🤓 |
End of preview. Expand
in Dataset Viewer.
README.md exists but content is empty.
- Downloads last month
- 34