Hacker News Reader

Comment by im_down_w_otp | Hacker News Reader

I love these breakdown writeups so much.

I'm also hoping that Intel puts out an Arc A770 class upgrade in their B-series line-up.

My workstation and my kids' playroom gaming computer both have A770's, and they've been really amazing for the price I paid, $269 and $190. My triple screen racing sim has an RX 7900 GRE ($499), and of the three the GRE has surprisingly been the least consistently stable (e.g. driver timeouts, crashes).

Granted, I came into the new Intel GPU game after they'd gone through 2 solid years of driver quality hell, but I've been really pleased with Intel's uncharacteristic focus and pace of improvement in both the hardware and especially the software. I really hope they keep it up.

View on HN

They won't make a B770 or C770 because they lose money on every card they sell. The prices are low because otherwise they would sell 0 and they already paid for the silicon. The intel graphics division is run by fools who won't give their cards a USP in the SR-IOV feature home labers have been asking for for years. Doing what AMD and Nvidia do but worse is not a profitable strategy. There's a 50% the whole division gets fired in the next year though.

>The intel graphics division is run by fools who won't give their cards a USP in the SR-IOV feature home labers have been asking for for years.

Intel are "fools" for not adding a feature that maybe a few thousand people care about?

A few thousand people is a lot more than the number of gamers who'd buy an Intel GPU at market price. If they don't raise their ASP into the black they're going to ax the whole division.

For reference the B580 die is nearly the size of the 4070 but sells for a third the price.

> For reference the B580 die is nearly the size of the 4070 but sells for a third the price.

Doesn't this suggest the B580 has worse yields? Die surface area isn't directly proportional to selling price.

Why would it say anything about yields?

What is says is that given the die area, Intel fails to capitalize on their chip relative to a 4070.

Intel GPUs are fabbed at TSMC so they're yields are the same as nvidia and amd.

mm2 is directly proportional to BOM cost and minimum profitable selling price. Intels software and hardware is much less efficient than nvidia so their products cost more to make and sell for less. Battlemage is at best break even, with alchemist they lost money. If they want to stay in business they need to raise their ASP. Since no one is buying intel for their games support they need to find a new market, thus SRIOV and local LLM. Unlike nvidia they don't need to worry about cannibalizing other business units.

Some counter points.

On the die size argument, which I see being echoed a lot online:

Why would a customer care or factor that into their purchasing decisions? Saying that these dGPUs with large dies are what is going to put Intel out of business is ludicrous and the Xe cores are shared amongst many of Intel's most lucrative products.

You can afford larger dies on N4 compared to when the 40-series were launched. It's no longer the leading edge node and yields have likely improved.

dGPUs have pretty expensive GDDR modules, I do not have data on the exact proportion but I would bet that the memory modules is the more important line item.

BoM matters less on lower volume (compared to mobile SoCs) dGPU units. Masks masks, R&D and validation are big fixed up-front costs.

Recurring software support is also independent of how many units get sold. Xe cores are shared by many Intel products (client & server CPUs, datacenter GPUs and gaming GPUs).

B580 is widely popular for gamers, Intel cannot keep up the demand at the moment. I doubt they need to unlock SRIOV on the gaming segment dGPUs to get rid of stock, as you seem to suggest. Their datacenter GPUs [1] offer support for SRIOV, as you probably already know, so I assume you are bemoaning market segmentation.

[1] https://www.youtube.com/watch?v=tLK_i-TQ3kQ -- Wendell's video on Flex 170 GPU from Intel - Subscription Free GPU Accelerated VDI on Proxmox 8.1

Gamers are buying at $250. If you pay more you can find B580 on the shelves now. That Xe cores and driver development are shared with other products is the only reason they haven't canceled the dGPU division. The hardware business is all about designing IP and amortizing the cost in as many products as possible.

Of course consumers don't care about die size and cost, only the value of the end product. The problem is intel and their negative margins. Maybe people are too blitz scale brained but running a negative margin hardware business is a really bad idea. Typically they target 50% margins.

Now if intel was healthy maybe they could survive a few negative margin products by subsidizing from their profitable SKUs. However intel is not healthy and reported losses every quarter last year. It's a sinking ship and they need a come to Jesus monent before they go bankrupt.

I'm arguing that their product segmentation strategy is stuck in the past and a significant reason why they're unprofitable. On desktop CPUs they lock ECC even though the Xeons are cheaper because they remove the E cores so AVX512 will work. On dGPUs they lock out SRIOV even though the desktop iGPUs support it. On the server they created accelerators which they then tried to charge for which means no one bothered to write integrations to make it work with their software. It's a broken culture which is so up it's own ass that management only cares about intel and is completely disconnected from the customers. Intel needs product features and value differentiators that aren't just negative margin products, changing product segmentation is low hanging fruit.

Being intransigent and the same as the competition but slightly worse had led them to unprofitability and soon, bankrupty. Interest rates are going up and companies carrying debt are going to be in a rough spot.

Intel needs to vastly streamline and cut down their SKUs (antithetical to the PM regime they have built up over the last 25+ years).

They need to stop artificially crippling their products to segment them. Simplify naming and have mid, great and greatest. The lowend shit products should be dumped. This means SRV-IO and ECC available everywhere. All of their GPUs should shift right by 8GB or more.

Streamlining their SKUs would streamline their wafer processing in the fabs.

SKU binning doesn't take any fab time.

They do have too many SKUs though.

How much has SKU explosion lead to mask set explosion and having to do runs, bank that and switching to another set of masks?

At this point they should just produce Xeons and support at most 2 sockets. I don't care if they fuse off features with cryptographic keys. Ship one piece of silicon to everyone, streamline the shit out of that.

ECC everywhere, no questions asked.

Basically none at all.

I haven't checked in a while but for client there's basically two die masks. In 13th and 14th gen half the SKUs were 12th gen masks.

Mobile is either the same as client or a smaller mobile efficiency die, but only one mobile mask.

For server there's just E cores and P core tiles. The limitation is packaging not fab.

The SKU is just fuses and written at validation after binning.

ECC everywhere would be a far better excuse for enterprises to upgrade their old PCs that 'AI PCs'

Agreed, Intel needs to win back customers. Removing the MBA-staple strategy of artificial market segmentation would be a strong signal that they are serious about becoming an actual customer-first company.

SR-IOV doesn't sell consumer cards, are you expecting Intel to produce an expensive XEON equivalent of Arc? I'd expect them to attempt capturing some LLM market share by loading up the cards with RAM rather than expending effort on niche features.

SR-IOV would sell more cards than Intel would be able to sell if they charged market rate for their GPUs. Same goes for not selling a local LLM high VRAM variant. Intel is just allergic to competing by offering a USP.

AMD Ryzen CPUs have ECC enabled but not officially supported. Intel still locks away the feature.

I have a couple of these too, and I strongly believe Intel is effectively subsidizing these to try to get a foothold in the market.

You get me equivalent of a $500 Nvidia card for around $300 or less. And it makes sense because Intel knows if they can get a foothold in this market they're that much more valuable to shareholders.

Great for gaming, no real downsides imo.

They should drop a $600 card with 128gb of vram. This is just barely possible without losses on every sale.

And then just watch heads explode.

At current market pricing on dramexchange, 128GB of 16Gbit GDDR6 chips would cost $499.58. That only leaves $100.42 for the PCB, GPU die, miscellaneous parts, manufacturing, packaging, shipping, the store’s margin, etcetera. I suspect that they could not do that without taking a loss.

I wonder if they could mix clamshell mode and quadrank to connect 64 memory chips to a GPU. If they connected 128GB of VRAM to a GPU, I would expect then to sell it for $2000, not $600.

Yup, just went with the $3 per GB formula.

GPU should be about $200 at TSMC (400-450mm2).

+ about $150 for the pcb, cooler and other stuff, I didn't consider

Times a 1.6 to 1.75 factor if they like actually being profitable (operations, rnd, sales, marketing, ...).

So about $1.5k, I guess.

Multiply that with a .33 "screw the competition" factor and my initial guess is almost spot on.

Real problem:

The largest GDDR7 package money can buy right now is 3GB. That's a 1376bit bus right there. GL fitting that to a sub 500mm2 die.

In the future you could put that amount of vram on a 512bit bus, tho.

Also normal DDR is getting really fast atm. 8 channel can already challenge most vram configurations. Maybe it's time soon to switch back to swappable memory.

>+ about $150 for the pcb, cooler and other stuff,

Assuming I had access to gerbers I could order replica of 5090 PCB for $65, including shipping. Intel PCB is half that. Again this is for a dude off the street buying 1-5 copies, not a bulk order.

This makes no sense. Currently 24GB is the sweet spot to stay competitive and 32GB is the maximum amount of memory you could stick on a card and still have a decent memory to bandwidth ratio.

What they should do instead is make the cards thinner and more efficient so that you can easily put two of them in a case.

They are definitely selling them at close to no profit, but they are not anywhere near subsidizing them unless they botched their supply chain so badly that they are overpaying the BOM costs.

R&D isn't free.

Even selling at cost is a subsidy.

I'm proud to support them. Intel is also selling their lunar lake chips fairly cheaply too. Let's all hope they make it through this rough patch. I can't imagine a world where we only have one x86 manufacturer.

> I can't imagine a world where we only have one x86 manufacturer.

Does it even matter? Some people won’t notice even if there are zero x86 manufacturers.

In fact I would say lots of people have not bought x86 CPU in while, between Mac, RPi and risc-v boards…

X86 is still needed for a lot of software. The emulation just isn't there yet.

That would be news to people on mac with Rosetta Stone / Crossover.

A lot of server code and specialized software won’t work.

Competition is always good

Server code tends to be extremely portable. Just recompile for the new architecture and you are done. The porting work for servers has already been done by the community. The main exception would be game servers, which are binary blobs, but box86/box64 can run those.

A number of businesses have switched to using arm EC2 servers from x86 EC2 servers for lower costs and things work fine on them.

Sometimes it can be ported, sometimes it can't.

x86 isn't going anywhere for a long time.

Imagine if AMD never existed. Everything about owning a computer would be worse. Likewise now we need Intel ( or someone else, perhaps a Chinese OEM) to make competition.

R&D is a sunk cost that is largely paid by their iGPUs. Selling at cost is not a subsidy and that is not relevant here since they should be making money off every sale. I tried estimating their costs a few months ago and found that they had room for up to a 10% margin on these, even after giving retailers a 10% margin. If they are not making money from these, it would be their fault for not building enough to leverage economics of scale.

https://slickdeals.net/f/17910114-acer-arc-a770-16gb-gpu-w-f...

> Acer Arc A770 16gb GPU w/Free Game & Shipping $229.99 $229.99 $399.99 at Newegg

Sure that margin is holding when they had to mark the first generation down to get them off the shelves. It would truly surprise me if they've made a significant profit off these cards.

I did a cost analysis for the B580, not the A770. I have no idea what their margins on the A770 look like.

Reply on news.ycombinator.com