Run on graphics cards....shortage
 

https://arstechnica.com/tech-policy/...card-shortage/

Amazing.........I was just ready to purchase a new machine and the processor problem appeared.....and now this

Pure insanity

Definitely a shortage of high end graphics cards.

But imo, the reason for the shortage isn't cryptocurrency mining.

I say that because I have a contact who works for a tech startup in Silicon Valley.

He tells me that maybe 3% to (at most) 5% of all high end graphics cards produced are being sold to cryptocurrency miners --

He also tells me the biggest market segment for high end graphics cards remains pc gaming --

And that the fastest growing market segment for high end graphics cards is actually tech startups like the one he works for.

They have a data center with as many servers as will fit in the floor space - and he says every server in there is stuffed to the gills with the newest Nvidia GPUs.

Fyi, the company he works for doesn't have a salable product yet.

He says in fact they may never have a salable product --

Because they are one of MANY startups funded by VC money trying to develop AI/deep learning --

And that nearly all the startups are in the same boat --

If it becomes apparent a startup has made some sort of breakthrough: That startup can expect more VC funding (a lot more in fact.)

Conversely, if it becomes apparent a startup has fallen too far behind the others: That startup can expect VC money to dry up.

All of that said, I feel ya --

Because I'm thinking about a machine with a high end graphics card too.



-jp

.

I've been hearing about mining and GPUs for quite a while now...well before this most recent article and shortage. I believe it is indeed a big factor...

I tried to find a figure for miners and the gpu market but could not . Everyone talks around it.

I do find 3-5% hard to believe.

Edited Transcript of NVDA earnings conference call or presentation 9-Nov-17 10:00pm GMT:
https://finance.yahoo.com/news/edite...054046030.html

Doing some math based on the transcript of Nvidia's most recent quarterly earnings conference call after close of market Nov 9, 2017:

For the quarter:

GPU revenue: $2.22 billion

GPU stands for Graphics Processing Unit (which is what we are talking about in this thread.)

Gaming revenue: $1.56 billion or approx 70.27% (calculated as 1.56/2.22 = 0.7027)

Data center revenue: $501 million or approx 22.57% (calculated as 0.501/2.22 = 0.22568)

Those two areas accounted for roughly 92.84% of Nvidia's GPU sales for the quarter.

That leaves approx 7.16% to be split among Nvidia's other/remaining market segments.

My understanding is that the two largest of these other/remaining market segments are GPU's sold to cryptocurrency miners followed by GPU's sold to the self driving car space.



All of that said, here's a link to an article that came out on Nov 10, 2017 (the day after Nvidia's most recent earnings call.)

Cryptocurrency Mining Sales Cool in Q3, Says Nvidia:
https://www.coindesk.com/cryptocurre...3-says-nvidia/

and:
If sales for cryptocurrency mining-related products for the quarter came in at $70 million --

And if total GPU sales for the quarter came in at $2.22 billion --

That means, for the quarter ended Sept 30, 2017, sales for cryptocurrency mining-related products were just 3.15% of total GPU sales.



-jp

.

Well then, if all that is accurate, then the stories about miners cleaning out the shelves is nonsense.

If that's case, why the shortage?

Makes no sense to me, but that happens a lot lately.

All the above was before bitcoin hit $20k....now that it's fallen back down, maybe the cards will return to the shelves...

Let's try this again...

Edited Transcript of NVDA earnings conference call or presentation 9-Nov-17 10:00pm GMT:
https://finance.yahoo.com/news/edite...054046030.html


Take a second look at the bolded part of the above quote.

The Data Center space is where sales of Nvidia GPUs is experiencing the most growth.

I think the reason this is so is because right now it's "race on" in the development of AI/deep learning arena -- and at least right now -- Nvidia GPUs are a big part of that.

I suspect the shortage of high end GPUs on the shelves is because demand for GPUs in the Data Center space for the quarter just ended (but not yet reported) may have caught Nvidia by surprise.

We'll know in a few weeks when they report earnings again.


-jp

.

deep learning artificial intelligence is the hottest sector right now. other company's that are involved in that type of chips are Qualcomm and Bosch. there must be some smaller companies as well that i don't know much about.

High End Video Cards in Servers
 

Why the heck are they putting high end expensive video cards in SERVERS?

I worked as an I.S. (information systems) manager for nine years. I would have been fired for doing that.

On many servers the monitor and video are not used very much.
That is a huge waste of money. Their CIO needs to review this. If he/she approved they need to be fired for wasting company money.

:confused:

The servers that are running AI/deep learning algorithms in the data centers I described (above) are configured to use a completely different info model than what you are describing --

Under a Standard Info Model --

• Just about everything is executed by the CPU - including requests from the server's operating system - plus requests from all of the users/visitors browsing all of the sites that are hosted on the server.

This info model works well if you are running standard or non-AI/deep learning programs on the server.

The programs being run are executing tasks such as verifying logon credentials, managing session state of site visitors, performing mathematical calculations, retrieving files, overwriting files, rendering requested web pages, reading data from databases, writing data to databases, and performing automated backups, etc.

Generally, this works rather well because the CPU is perfectly capable of handling all of the requests being made.

But this standard info model doesn't work so well if the number of requests rises beyond a certain point. If that happens - and a good example might be a DOS or denial of service attack: The CPU reaches a point where it simply cannot keep up with all of the new requests coming in -- and the server slows to a crawl.

This standard info model also doesn't work so well if the server's primary purpose is to run AI/deep learning algorithms.

Don't get me wrong - under a standard info model where everything is executed by the CPU:

The server is perfectly capable of running AI/deep learning algorithms -- just not quickly.


Why use High End GPUs? --

• High end GPUs are faster and more powerful than a CPU. They are able to process a greater number of decisions/transactions per second than a CPU.

• The AI/deep learning algorithms researchers are working on are often called upon to process a LOT more transactions/calculations before reaching a simple classification decision (such as 0 or 1) than non-AI/deep learning algorithms.

For these two reasons, pretty early on, researchers working on AI discovered that GPUs are better than CPUs for running AI/deep learning algorithms.

Once chip makers like AMD, Intel, Nvidia, and others realized that - they began working on GPUs designed specifically for AI.

Somewhere along the way - Nvidia began snatching up market share from AMD and Intel in this new space -- Imo because they've been cranking out GPUs with specs that are far and away superior to those made by AMD and Intel. (That doesn't mean AMD or Intel or someone else couldn't announce a major breakthrough tomorrow. But right now as I type this Nvidia is the clear leader in this space.)



Alternate Info Model --

The servers in the data centers I described in my previous post (above) are configured in such a way that nearly all of the requests being made by the AI/deep learning algorithms are executed by banks of High End GPUs - instead of by the CPU.

As a result the AI/deep learning algorithms are running at max possible speed (up to the limits imposed by the latest available technology.)

Hope I managed to type most of that out in a way that makes sense,



-jp

.

Clearly the GPUs are not being used for video. GPUs are being used because they are better at certain math processes than a CPU. In the case of cryptocurrency mining, GPUs are on the order of 80x faster than a CPU when is comes to solving the transaction blocks. For bitcoin mining, an ASIC miner is better. But other cryptocurrencies (like Ethereum) are processed more quickly using GPUs. So depending on the algorithm, a GPU can be much better suited to the task than a CPU.

Ding!

Good stuff from Jeff too👍

Don't know why anyone would think that the crypto craze isn't driving a lot of the sales.

Saw this report today and hearing of an AMD card (the Radeon 56) listing for $ 399, selling for $999+ (at retail, not just ebay) and flying off the shelf.

These cards are better investments than the sh*tcoins




Analysts suggested the popularity of mining for ether, the cryptocurrency running on the Ethereum network, drove sales of graphics chips from AMD and Nvidia Corp. NVDA, more than the companies were letting on in the fourth quarter.