nVidia Fermi aka GT300 under investigation!

In the name of truth!
source: semiaccurate.com

Nvidia fakes Fermi boards at GPU Technology Conference

Jen-Hsun says one thing, pictures prove another

by Charlie Demerjian

October 1, 2009

WHAT DO YOU do when have a major conference planned to introduce a card, but you don't have a card? You fake it, and Nvidia did just that.

In a really pathetic display, Nvidia actually faked the introduction of its latest video card, because it simply doesn't have boards to show. Why? Because it didn't get enough parts to properly bring them up, much less make demo boards. Why do we say they are faked? If you look at the pictures, it is painfully obvious that Fermi cards don't exist. Well, painful if you happen to be Dear Leader who waved fakes around and hopes to get away with it, but hilarious if you are anyone not working at Nvidia.

Lets go through this in pictures, shall we? Please note, none of the pictures are ours, and since you can find them on many other sites, they are likely Nvidia PR material. Since the company seems to have forgotten to mail us a copy, we will point to the ones at PC Watch.

Fermi chip heat spreader closeup

The first one is from the always great PC Watch, specifically here, but we cropped it a bit for size. Some things to note on it. First is that the second set of digits on the first line says 0935A1. A1 is for first silicon, something that, when coupled with a direct quote from Jen-Hsun of, "You are currently looking at a few day old silicon" (from here select the video "See video of Jen-Hsun Huang announcing Fermi") kind of blows the whole 'silicon in Santa Clara last spring' story out of the water. I wonder if anyone will retract that, or just change the article retroactively?

To make matters worse, the other digits are a date code, 0935 means 2009, work week 35. If Nvidia starts its work week on the first full week of the year like everyone else, that would put WW35 at August 28 to September 5, 2009. Where have I heard that date before? Oh yeah, here.

Last but not least, you have the funniest part, the number seven hand written on the GPU. While this isn't proof of anything, normally when you have small lots of samples, you write numbers on them to identify which one is which as an aid to debugging. Where did we hear that number seven again? Oh yeah, here again.

Here is a free tip to Nvidia PR, when you go after people who are right to try and undermine their credibility, at least be correct, or close to it. You were neither this time. Another tip, coach your messengers better, or at least have them read the article you are attacking. Poor showing guys. Did Intel hire away your best attack dogs along with your engineers?

'Fermi' end plate closeup

The next one is also from PC Watch, specifically here, again cropped to emphasize detail. Notice how the end plate vents on the second slot are blocked off? That gives a lot of cooling, eh? You really want to have early samples well cooled so you get good data off of them, and also so you don't damage 14.3% of your chips at once.

Then there are the screws. Notice that the three screws that hold the end plate on are, well, generic wood screws. Large flat head phillips screws. Home Depot grade screws that don't even sit flush. If a card is real, you hold it on with the bolts on either side of the DVI connector. Go look at any GPU you have, do you see wood screws that don't mount flush or DVI flanking bolts?

5870 on top of a 4890, both real

As you can see from the picture above of my Radeon HD5870 sitting on my HD4890, the end plate is more than capable of being held on without wood screws. It mounts to the PCB via a bolt or two, sometimes solder, and the soldered down DVI pins that go through the PCB VIAs.

The back of the 'Fermi' endplate

If you look at the back of the fake Fermi, above cropped from PC Watch here, you can see that the expected DVI connector wires are not there, just solder filled holes. No stubs, no tool marks from where they would be cut out. Basically, the DVI port isn't connected to anything with solder, so they had to use screws on the plate.

Compare and contrast this to the 4890 and 5870 above, which, unlike the Fermi, are real. Also, note that the second SLI connector is half covered by the shroud. I wonder how much bandwidth that adds? Nvidia's comedy team strikes again, as long as your initials are not JHH. If they are, you look pretty silly in front of thousands of people. In fact, if you look at the video, you can see the wood screws. You can also see them in the well watermarked photo on Fudzilla here, but I doubt Nvidia will be so happy when it realizes that the word is out.

If you look at the back of the card below, taken from the same PC Watch pic as the last one, again cropped, the hilarity continues. The card has two power connectors, one on the end, the other on the top of the card, bottom of the picture.

'Fermi' power connectors and back of the board

Look closely, the 8-pin connector on the end doesn't line up with the 8-pin solder joints, they are 90 degrees off, and there are two extra heavy duty solder through holes right up to the back edge of the PCB. Those lead to... well, not the power connector. PCIe specs make the pins on the 8-pin connector far too deep for it to be them. In fact, it can't really run to anything, since the 8-pin connectors would block it.

The 6-pin connector, on the other hand, lines up with, umm, nothing. There is a potential 4-pin floppy/sound/jumper block below it, but you can clearly see there is nothing in the vias, not even solder. The 6-pin connector connects to nothing, and nothing is holding it in. Except glue. Notice the connector is black and the hole below it shows white. The only real question now is, Elmers or glue stick?

To make matters worse, the mounting holes for the 8-pin connector, which should be between the 6-pin and 8-pin fakes if the card was real, are empty. Piss-poor fake job guys. Go read your fanboi forums, they do a better job, and work for much cheaper than your 'geniuses'.

The two solder connections under the 8-pin connector on the end of the card are likely connected to a 6-pin connector that normally is next to where the 8-pin on top was before it was removed. The problem here is that there are two solder holes, four are missing, almost like they ran off the end of the card and were chopped off.

But that would mean the PCB was crudely cut off with base power tools, not even proper PCB making equipment, in order to fake Fermi. If you don't even look all that closely, you will see two barcode stickers on the bottom side of the card end. They are literally cut in half, about what you would expect if Dear Leader came down and ordered a crude fake. That is unfair, he would never order workers to make a crude fake, he would order a high quality fake, and Nvidia engineering would deliver a crude one.

You should also note that the mounting holes for the 6-pin connector are so close to the edge of the PCB that they would offer no real mechanical support and simply crack if used. No engineer would do that, they would make a longer board. And they did. And Nvidia chopped it off to fake Fermi. Several other bits look oddly truncated as well.

The chips arrived when I said they did, in the number I said they did. Nvidia didn't like this, so it used Fudo, who I honestly have the highest respect for, in a way that trashed his reputation. It has been trying to do this for a long time, and sadly, it looks to be getting close to success.

The board has wood screws crudely driven through it. The vents on the end plate are blocked. The DVI connector is not soldered to anything, The SLI connectors are somewhat covered by a heat shield. The 8-pin power connector is connected to nothing. The 6-pin connector is connected to the PCB with glue, not pins and solder. The board is crudely chopped off with power tools. The 8-pin connector that should be there is not. The 6-pin connector that should be there is cut. The mounting holes are too close to the edge. There are also likely many more flaws, but this should be enough to prove a point.

In the end, what you have here is a faked Fermi board. Jen-Hsun held up a scam card. If you watch the video here, he says, "This puppy here, is Fermi". Bullshit.S|A

Note 1: Nvidia PR was asked to comment on the faked cards earlier this evening. Their reply was, "I'm glad you're asking us before you write. That statement is false. The product that we displayed was an actual Fermi board. The demo ran on Fermi silicon." We do not believe all of that statement.

Note 2: Thanks to readers Michael and Bill for emailing about this while it was being written.

Note 3: Thanks to the forum denziens, starting here, who are already on top of this story.

Note 4: Thanks to PC Watch for putting up such great pictures.

Note 5: Thanks to Nvidia for providing such great source material for articles.

source: jasoncross.org

A Few Thoughts on Nvidia’s Fermi

Today was the start of Nvidia’s GPU Technology Conference. It’s really still just the NVISION conference, because it’s not much of a “industry-wide” conference if ATI and Intel aren’t there. The biggest announcement of the show is undoubtedly the unveiling of Nvidia’s next-generation GPU, code-named Fermi. I’m not sure why they named the chip after Enrico Fermi, who is best known for his work with radioactive substances and controlled nuclear reactions and stuff. But as code-names go, physicists are cool, so I’ll let it slide.

I won’t bother to summarize all the individual features that were revealed today. Tech Report has a excellent article on it, so does AnandTech. I’m just going to editorialize a bit with some of my thoughts based on what we know (and don’t know) so far.

First, boards based on Fermi are going to cost a considerable bit more than the Radeon HD 5870 and 5850, which are ATI’s competing DX11 cards that just launched. The RV870 GPU powering ATI’s cards is 334 mm². It has a 256-bit memory interface. Nvidia didn’t talk about GPU size, but it did say that Fermi is 3.0 billion transistors – 40% bigger than RV870’s 2.15 billion. So, figure a chip somewhere around the 460-480 mm² mark. That’s huge.

The chip being 40% bigger doesn’t mean 40% more expensive to produce, though. Imagine chips A and B. Both are 40nm chips made at TSMC. Chip A can fit 100 chips on a wafer, and Chip B can fit 60 chips on a wafer, because it’s 40% bigger. But as chip size grows, it’s harder for the whole chip to come out without flaws, so the yields are worse. Chip A has a yield of 75% – three-fourths of all the chips on the wafer function properly within the intended specs. Chip B has a yield of 60%, because it’s so much larger. That means you’ll get 75 good chips on a wafer for Chip A, but 36 good chips for Chip B. That’s less than half.

In other words, depending on how the yield situation works out, Fermi could be twice as expensive to produce as RV870. Hell, it could be worse. We really have no way of knowing, except to say that a 40% larger chip is usually well more than 40% more expensive to make.

It’s not just the chip, either. A 384-bit memory interface means Fermi-based cards will likely have either 768 MB (not likely) or 1.5 GB of RAM, so that’s higher RAM costs. It also means more PCB layers on the board itself. So aside from higher chip costs, the board costs of Fermi-based products will be higher than Radeon 5800 products.

So if Fermi-based products are going to be considerably more expensive than Radeon 5800 products, what about performance? Well, all Nvidia has talked about so far are the chip design elements that impact GPU compute, rather than traditional graphics. There’s quite a lot there. Nvidia has clearly spent a fair chunk of the transistor budget doing things like dramatically improving double-precision floating point performance, increasing cache sizes, ECC memory support, and so on. These things typically do nothing at all for typical graphics performance (games and stuff). So the chip is 40% more transistors, but that won’t necessarily translate into 40% higher frame rates.

Nvidia seems to be gearing the world up for this. The mantra they keep chanting is that “graphics performance isn’t enough anymore.” Compute really matters a whole heckuva lot, they tell us. This sounds like PR code for “the card is going to be 50% more expensive than the competition and not 50% faster in games, so please place as much importance on GPU compute apps as possible so we look like a better value.”

You know how drill sergeants tell recruits to begin and end everything they say with “sir?” Sir, yes sir! Sir, I didn’t mean to shoot the sergeants toe off, sir! That’s what it’s like listening to Nvidia these days, only with “CUDA” instead of “Sir.” For over a year, Nvidia has told everyone who will listen that GPU compute is super duper important, and has very aggressively flogged PhysX and CUDA. And you know what? Consumers just don’t care all that much. Maybe one day, when there are robust standards and quite a few GPU-accelerated applications that normal people use all the time, the average consumer will want a graphics card to make its non-gaming apps go faster just as much as it wants it to make its games go faster and look better. But we’re not there yet, and we’re not going to be there in the next six months, as much as Nvidia would like us to be.

So Nvidia’s facing a tough sell in Q1 2010 (or maybe late 2009) when the first Fermi-based cards go on sale. They’ll almost certainly cost $399 or more, judging by what we know so far. ATI has a chip and board design that will let them push Radeon HD 5850 cards below $200 and 5870 cards below $250 within the next six months, if they want to. Is a modest increase in frame rate and much higher performance in GPU compute apps going to be worth such a broad difference in price? Will it be a moot point, because the cards will be out of the cost and power budget for most consumers (and OEMs)?

UPDATE: Nvidia confirms
source Fudzilla.com
Fermi that was running PhysX at Jensen's GTC keynote was real but the one that we all took pictures of was a mock-up.

The real engineering sample card is full of dangling wires. Basically, it looks like an octopus of modules and wires and the top brass at Nvidia didn't want to give that thing to Jensen. This was confirmed today by a top Nvidia VP president class chap, but the same person did confirm that the card is real and that it does exist.

Charlie was surely right about this one here. The card that Jensen showed and we all pictured is not a working sample. Some key people at Nvidia have promised to show us the picture of the real one. Stay tuned.