The OFFICIAL tech stuff thread

Kilemall

@Hog said in The OFFICIAL tech stuff thread:

So it mostly seems to be about utilizing otherwise wasted PCIe lanes and not something that the graphics card itself would use directly.

That’s weird, I would have thought it was as used to preload software off the channels for the use of the card. Those things are practically standalone computers anyway.

Or maybe dump the output of the video card for AI work.

Gators1

The AI stuff could use disk cache. I have run out of memory trying to upscale my pornz.

Kilemall

@Gators1 said in The OFFICIAL tech stuff thread:

The AI stuff could use disk cache. I have run out of memory trying to upscale my pornz.

Ooooh ya, the very thing you were on about Hog- home compiling of upscale drivers and the like.

If they truly have run out of Moore’s law, maybe we’ll see more crazy architecture to get around it.

RISC-V Week: 7 days only using RISC-V computers

RISC-V has come along a lot further than I imagined it could in a short time. The above guy used a couple of cheap, raspberry PI-like, RISC-V computers for a whole week and surprised me with how much he could do with them. ARM still have all their smarts and skills and will remain in demand for mid to high range phones and other devices that require it but I think RISC-V is shaping up to take a sizeable chunk of their lower end and more generic device market.

tigger

Whatever he was doing, it was blinking a lot.

Tazz

Hence the 503 window. ;)

Lob12

Tigger just admitted that this award winning forum is coded by child labour.

Great.

Gators1

@tigger Interesting approach to raising your kid. You are turning him into a nerd that nobody will want to date, but at the same time giving him the skills to build his own girlfriend.

tigger

@Gators1 said in The OFFICIAL tech stuff thread:

@tigger Interesting approach to raising your kid. You are turning him into a nerd that nobody will want to date, but at the same time giving him the skills to build his own girlfriend.

What’s actually kind of cool is that the nerds seem to be also the popular ones and the ones that do all the sports.

Gators1

True, being a nerd these days isn’t want it used to be. I guess maybe they are even getting laid by womenz trying to get a piece of their tech startup IPO.

tigger

I guess I should teach him that there are cases when a crosscheck is the appropriate solution to your problems, but otherwise he’s all good.

Testing out the Milk-V Duo - The new $9 RISC-V eSBC that runs Linux!

The above is another cheap RISC-V single board computer except this one is only $9. Then at 21:08 he says, “Oh I forgot to mention, this one has a tensor processing unit. I don’t know how good it is but it’s designed to do some sort of AI.”

Nine dollars, fuck me.

Tazz

It probably gains access to everything you own and will shut everyone down in the blink of an eye.

https://twitter.com/pdrmnvd/status/1707395736458207430?s=20

@Kilemall said in The OFFICIAL tech stuff thread:

That’s tantamount to fraud.

I guess we can’t tell from that one email if his request went anywhere. Given the tension between attracting/retaining users with a useful product and wanting to squeeze every last dollar out of them, I guess it shouldn’t be surprising that some people within google would be pushing for what he’s asking for.

I also don’t know who he is in terms of seniority but the guy he’s replying to “Anil Sabharwal” seems to be a big deal from my brief search.

Facebook appears to be on an interesting journey to destroy what made them popular in the first place. Well, they kind of already had but has anyone else been inundated with friend requests lately? I’ve gone from 500 to 1K friends in a week and. several times now when I’ve gone to confirm some of the new requests it says I can’t because the person who requested it has maxed out at 5K friends. If Facebook is behind the change, it seems they are trying to make the whole “friend” thing even more meaningless than it was. Maybe as a prelude to turning it off and making everything public? Dunno.

tigger

@Hog said in The OFFICIAL tech stuff thread:

@Kilemall said in The OFFICIAL tech stuff thread:

That’s tantamount to fraud.

I guess we can’t tell from that one email if his request went anywhere. Given the tension between attracting/retaining users with a useful product and wanting to squeeze every last dollar out of them, I guess it shouldn’t be surprising that some people within google would be pushing for what he’s asking for.

I also don’t know who he is in terms of seniority but the guy he’s replying to “Anil Sabharwal” seems to be a big deal from my brief search.

What is the request though? As far as I can tell it’s something to do with rolling back a launch which redirected some queries from chrome away from search, e.g. if you type “Facebook” going directly to Facebook instead of Google. I’m unsure I see how that is “fraud”.

Kilemall

@tigger said in The OFFICIAL tech stuff thread:

@Hog said in The OFFICIAL tech stuff thread:

@Kilemall said in The OFFICIAL tech stuff thread:

That’s tantamount to fraud.

I guess we can’t tell from that one email if his request went anywhere. Given the tension between attracting/retaining users with a useful product and wanting to squeeze every last dollar out of them, I guess it shouldn’t be surprising that some people within google would be pushing for what he’s asking for.

I also don’t know who he is in terms of seniority but the guy he’s replying to “Anil Sabharwal” seems to be a big deal from my brief search.

What is the request though? As far as I can tell it’s something to do with rolling back a launch which redirected some queries from chrome away from search, e.g. if you type “Facebook” going directly to Facebook instead of Google. I’m unsure I see how that is “fraud”.

It’s fraud because the people paying for the ad clicks are getting falsely inflated results by a manipulation of the search results from a particular browser rather than actively interested quality clicks. Conversion rate to paying customer is going to be lower, so value paid is not value bought.

tigger

@Kilemall said in The OFFICIAL tech stuff thread:

@tigger said in The OFFICIAL tech stuff thread:

@Hog said in The OFFICIAL tech stuff thread:

@Kilemall said in The OFFICIAL tech stuff thread:

That’s tantamount to fraud.

I guess we can’t tell from that one email if his request went anywhere. Given the tension between attracting/retaining users with a useful product and wanting to squeeze every last dollar out of them, I guess it shouldn’t be surprising that some people within google would be pushing for what he’s asking for.

I also don’t know who he is in terms of seniority but the guy he’s replying to “Anil Sabharwal” seems to be a big deal from my brief search.

What is the request though? As far as I can tell it’s something to do with rolling back a launch which redirected some queries from chrome away from search, e.g. if you type “Facebook” going directly to Facebook instead of Google. I’m unsure I see how that is “fraud”.

It’s fraud because the people paying for the ad clicks are getting falsely inflated results by a manipulation of the search results from a particular browser rather than actively interested quality clicks. Conversion rate to paying customer is going to be lower, so value paid is not value bought.

Maybe explain what you believe is happening and what evidence the email provides? Because I am not following you at all. Everyone in the thread is, as far as I can tell, suggesting the exact opposite.

Jam

I Googled SQV related to sales (try just “SQV” without sales and see what you get) and I got, in the “headline”"

Sales Quantity Variance (SQV) Difference between the actual contribution margin and the standard contribution with sales mix and sales volume held constant. Difference between the actual product mix and the standard product mix with sales volume and contribution margin held constant.

And then I searched under “query rollback”, a new term for me and got:

The query executes a ROLLBACK TRANSACTION statement. This statement abandons all changes made inside the transaction. The query ends before reaching either of these two statements. In that case, BigQuery automatically rolls back the transaction.

If I had a very cynical mind (a condition not contrary to fact) I’d be thinking that, a query rollback on a Google search related to a product search, would end in a shorter or shallower response and allow selected ads to pop up quicker in the responses?

Would this amount to ‘putting one’s finger on the scales’ a bit or some form of query/response results manipulation?

I don’t pretend to understand all the jargon and don’t want to dig much deeper, but it seems like the search/query team and the sales of advertising team may be considering working together to be sure that the peeps buying ads get more eyeball time at the expense of more “honest” responses to queries?

As I’ve said, I do not care enough to try and dig more deeply and I can be misinterpreting what I read due to a not so deep understanding of the terms.