China customs seize thousands of GPUs from XFX

mongeese

Posts: 643   +123
Staff
In brief: This week, Chinese authorities seized a shipment of XFX GPUs at a port on the border between Hong Kong and Shenzhen. It contained 5,840 GPUs that local media identified as mostly AMD Radeon RX 6700 XTs headed for XFX's Chinese sales partners.

According to a report from Chinese customs, the shipment was seized after a routine inspection found three mislabelled GPUs inside. A subsequent investigation discovered that the company shipping them had declared them all with the wrong names and specifications.

Officials say the shipment's value exceeds ¥20 million ($3.15 million).

Authorities did not mention XFX by name, but a video posted to Weibo on Saturday shows officials opening the shipment where the XFX logo is clearly seen. The Chinese company's website is currently down, and Alibaba-run storefront Tmall has stopped listing its products. China-based online retailer JD continues to sell them.

Although owned by Hong Kong-based Pine Technology Holdings, XFX is headquartered in California. The hardware company's operations outside of China have continued as normal.

It's required to pay import taxes when shipping products to China, which might be what XFX was trying to avoid by mislabeling its GPUs. According to the Chinese outlet My Drivers, the practice is common among OEMs that ship to China.

Permalink to story.

 
You have to be an utter ***** to try to cheat taxing authorities, especially for US company, in current political climate vs China. Seems greed is hard to resist.

In truth I could kind of understand if it was a one box, but I bet that was at least whole container of boxes.
 
Here's why AMD GPU's of the 6K series can't be found anywhere on the Steam Survey. AMD's partners sell directly to Miners by the truck load.

There was also a you tube vid from China with thousands of Radeon 6K GPUs just received by a dealer over there who was advertising the fact to potential Miner buyers.

He was gonna sell them by the palette only.
 
You have to be an utter ***** to try to cheat taxing authorities, especially for US company, in current political climate vs China. Seems greed is hard to resist.

In truth I could kind of understand if it was a one box, but I bet that was at least whole container of boxes.
That’s not how it works... you need to bribe the right person (or people) when attempting this.... I guess they didn’t bribe the right people - or they were sold out to a higher bidder...
 
I don't buy it. Well I do but I don't: I believe XFX are just like any other GPU manufacturers (That is, scum) but I don't believe they'd open themselves up by getting the CCP after them: they're extremely likely using a distributor. Maybe a legit one, maybe just someone's cousin's name in the company to get the necessary legal plausible deniability but I don't think it's XFX directly.

We know this because we know that's how they sell full pallets of GPUs to miners or scalp them on ebay, in fact we already know MSI was dumb enough to do so with an employee directly connected to them but the rest of them do it to, just with more obfuscation.
 
Sadly, that survey is beyond unreliable. It has been said and proven over and over.
If you mean that devices running Steam <> all devices, well, of course.

If you mean that it's unreliable as far as devices with Steam on them, then I have yet to hear that proof. Do you have a link?

Personally I'm of the belief that the Steam Hardware Survey is a reasonable overview of the hardware in use by Steam users, and also that Steam users are a reasonable proxy for PC gamers generally.

Of course I agree that the GPU market includes both PC gamers and miners, and that miners are mostly not running Steam nor are represented on the survey.
 
Over the years I've purchased two XFX cards. Both failed within a month. The first one I was able to return to NewEgg without issue. The second card, Kyle from HardOCP helped me get my money back after XFX gave me the run around. Couldn't happen to a more deserving company, imo. I'm sure some people like XFX, but for me, that name just brings up bad memories and lots of headache.
 
If you mean that devices running Steam <> all devices, well, of course.

If you mean that it's unreliable as far as devices with Steam on them, then I have yet to hear that proof. Do you have a link?

Personally I'm of the belief that the Steam Hardware Survey is a reasonable overview of the hardware in use by Steam users, and also that Steam users are a reasonable proxy for PC gamers generally.

Of course I agree that the GPU market includes both PC gamers and miners, and that miners are mostly not running Steam nor are represented on the survey.
It would had taken you less time doing a google search than typing all that just to prove your bias.
 
Yeah, I googled for Steam Survey Unreliable. Here's Google's pick for top answer which is a hardly a rebuke to a survey being taken as a survey:

"The standard disclaimer applies, in that Steam's survey results are not a completely accurate snapshot of market share, or even totally representative of the entire Steam population. However, it is one of the best resources we have to observe trends in hardware use among gamers (like Linux taking a 1% OS share)."

Other than corner-case issues involving "my version of hardware X is counted as alternate flavor Y" I did not see a single result indicating why the steam survey results would not be a reasonable estimate of steam user's hardware, which is why I asked you for this "proof" you claim to have seen.

And what "bias" are you even talking about. Is this like some weird AMD vs Nvidia issue I've stumbled into in your head?
 
Over the years I've purchased two XFX cards. Both failed within a month. The first one I was able to return to NewEgg without issue. The second card, Kyle from HardOCP helped me get my money back after XFX gave me the run around. Couldn't happen to a more deserving company, imo. I'm sure some people like XFX, but for me, that name just brings up bad memories and lots of headache.

Same here. Their products are junk.
 
The 6700xt would be perfect if there were enough of them to get under MSRP. How much you wanna bet these cards will end up destroyed somehow?
Sadly, that survey is beyond unreliable. It has been said and proven over and over.
The thing is, every time somebody claims this, they give 0 sources and 0 arguments to back up their argument, nor any better sources for this information.
It would had taken you less time doing a google search than typing all that just to prove your bias.
Oh hey look. Funny how history repeats itself....
Over the years I've purchased two XFX cards. Both failed within a month. The first one I was able to return to NewEgg without issue. The second card, Kyle from HardOCP helped me get my money back after XFX gave me the run around. Couldn't happen to a more deserving company, imo. I'm sure some people like XFX, but for me, that name just brings up bad memories and lots of headache.
My XFX 480 GTR black is still running strong 6 years later, running at the OC limit most 480s hit.

Both my powercolor vega 64s have kicked the bucket in spectacular fashion.

Any brand can have some lemons.
 
Here's why AMD GPU's of the 6K series can't be found anywhere on the Steam Survey. AMD's partners sell directly to Miners by the truck load.

There was also a you tube vid from China with thousands of Radeon 6K GPUs just received by a dealer over there who was advertising the fact to potential Miner buyers.

He was gonna sell them by the palette only.
If you think about the fact that when mining is thriving, and you can’t find any GPUs for sale, then you will know that AMD, Nvidia or their board partners are selling GPUs from their back doors to miners. LHR or not doesn’t matter because you can hardly find any GPU, regardless whether it’s FHR or LHR. How do you think miners are showing off massive mining farm full of Nvidia GPUs?
 
Yeah, I googled for Steam Survey Unreliable. Here's Google's pick for top answer which is a hardly a rebuke to a survey being taken as a survey:

"The standard disclaimer applies, in that Steam's survey results are not a completely accurate snapshot of market share, or even totally representative of the entire Steam population. However, it is one of the best resources we have to observe trends in hardware use among gamers (like Linux taking a 1% OS share)."

Other than corner-case issues involving "my version of hardware X is counted as alternate flavor Y" I did not see a single result indicating why the steam survey results would not be a reasonable estimate of steam user's hardware, which is why I asked you for this "proof" you claim to have seen.

And what "bias" are you even talking about. Is this like some weird AMD vs Nvidia issue I've stumbled into in your head?
Then you have your results as you want them to be.

But fine, you got a minute of my time. Funny that every time something similar is posted, myself and others provide their answers (in over 15 years in steam, I never got a survey when I had a radeon, plenty of times when on a geforce and plenty of others have the same experience) to results dating back more than 5 years (google it, not doing it again for you)

Fine …again…are you truly that dense in assuming that every single amd gpu sold, especially the 6K serias has somehow magically ended in system that steam is not installed?

Anyways, take it as you wish, I wasted enough time on this nonsense….which will return on the next article about the unreliable steam hardware survey.
 
The thing is, every time somebody claims this, they give 0 sources and 0
And since every time this bs is brought up, myself and others have provided links and proof. So that tells me that either you dont read the posts or you are one of those that simply need to troll others with the same false insinuations.
 
OK, I'm at least understanding your hypothesis, which appears to be that the survey results undercount installed AMD hardware.

My first question would be, is this an assertion that say AMD is also making? Given they are a major supplier to Valve they'd certainly be in a position to have any concerns heard. Even if they weren't they have enough of a pulpit in the kinds of publications we all follow that they could get a message out.

My second would be, do the results in the Steam survey differ materially from other, non-Valve sourced, reports re: market share?

This is a new issue to me so I can't claim to have done exhaustive research but trying to quickly triangulate now from other sources, I don't see a big discrepancy there. (it's a little tricky because most results are point-in-time sales for say one quarter, while survey results are installed base which reacts slowly.)

And finally, what would Valve's motivation be? I feel like the primary value of this report for developers in understanding what level of performance is out there in the game-buying audience, with the supplier of that performance being basically irrelevant (unless you're willing to ignore one or the other, which no developer is.)

I do get that based on your personal experience getting invites on one card and not the other may leave you with this question, but of course it is a random sample and the experiences of a single individual are pretty much just random noise.
 
OK, I'm at least understanding your hypothesis, which appears to be that the survey results undercount installed AMD hardware.

My first question would be, is this an assertion that say AMD is also making? Given they are a major supplier to Valve they'd certainly be in a position to have any concerns heard. Even if they weren't they have enough of a pulpit in the kinds of publications we all follow that they could get a message out.

My second would be, do the results in the Steam survey differ materially from other, non-Valve sourced, reports re: market share?

This is a new issue to me so I can't claim to have done exhaustive research but trying to quickly triangulate now from other sources, I don't see a big discrepancy there. (it's a little tricky because most results are point-in-time sales for say one quarter, while survey results are installed base which reacts slowly.)

And finally, what would Valve's motivation be? I feel like the primary value of this report for developers in understanding what level of performance is out there in the game-buying audience, with the supplier of that performance being basically irrelevant (unless you're willing to ignore one or the other, which no developer is.)

I do get that based on your personal experience getting invites on one card and not the other may leave you with this question, but of course it is a random sample and the experiences of a single individual are pretty much just random noise.

On mobile, so hard to properly format.

I honestly don’t know the reason for this and I dont think that valve is doing something nefarious.

But it ha been discussed before because it is not a new issue.

What is worse, yes, we simply dont have anything even close to the SHS, but something in it is broken and has been for a while.

Just look thru the previous post in this site about how weird and inconsistent the surveys are.

There is a way for users to submit their data without receiving a survey, but very few know about it or do it that way.

Personally, if I put the SHS on one side and the lack of AMD gpus available to buy due to being sold out all the time, then who the hell is buying all of them and hiding them from valve?

And we cannot say only miners, because rdna and rdna2 are not desirable to miners.
 
Back