Is there enough volume of this chip for it to mean anything to anyone? If there are only a few that no one can’t buy, and they are only better than nvidia at 1 out of 10’000 workloads, can people even use their existence to try to get lower H100 prices?
I am currently analyzing the feasibility and economics of building a cloud service for AI training. This involves "day to day" analysis changes since all the "parts" are moving.
I use Android (work) and iPhone (personal) phones.
My experience with Apps on iPhone is much better. The Apps are better, there is less trash, the Apps get less of my personal data.
This is valuable to me.
Maybe App developers for iPhone are just better human beings than Android developers. I think that’s not the case, and it’s due to the walled garden.
We can argue about why this is the case. But at my company many have both ecosystems in their pockets all day, and no one has a private Android phone for these reason. Sample size 100+ people, with better things to do in life than recompiling code for their phones.
I get that if you want to hack on your phone, iPhone is not the right phone for you. The user experience for hacking on your phone vs using your phone as a non SW developer are allowed to be completely different. One market is significantly larger than the other.
The federal government foreign policy is giving full bright scholarships to the best of the best Chinese students so that they get PhDs fully paid at the best US tech universities. And then kicking them out for loosing a lottery 3 times and sending them back to China so that they have no choice but working in products and companies that directly compete to the US.
Selling or not selling tech to China is the least of the US problems.
There I read that Intel Gaudi is behind all NVIDIA products in all metrics except performance per dollar.
Since one can only access these in the Intel cloud, where Intel has essentially made the few of these that exist freely available, isn’t perf per dollar kind of a joke metric? I mean, if they are free to use, they have infinite perf per dollar. But if Intel were to sell them, you were to buy them, and then you were to pay for their power, then wouldn’t perf per dollar look completely different ?
The article seems to try really hard to avoid saying how much perf per watt these have and how expensive they are. Reads a bit like a marketing piece as part of some Intel / data bricks collab (all companies and vendors do these).
> wouldn’t perf per dollar look completely different
Isn't the cost of actually running the card dwarfed by the cost of the card itself? e.g. even if we look at an old card like the A100 at 100% 24/7 that's about 3500 kWh per year which is what $300-$500 depending on the location?
if we Intel could come out with a card that could do anything the A100 does at half the price but 4x power usage it could still be somewhat competitive.
That’d depend on how many of these cards you need to get the same perf as an A100. Most of the results being shown are batch 1 (use the cards to serve 1 user). But in practice you’d use a single A100 to serve thousands of users concurrently (you can’t do that with the Gaudi 2 though).
The article measures that Gaudi 2 is competitive at “latency for batch size 1”, but that isn’t really a metric anyone cares about.
So Intel would need to sell these for much less than half, and then perf per watt would need to be much better (the article measures that it is currently worse than A100).
Comparing these things is hard, “if Intel were to sell these” is already speculation, since they aren’t on sale.
The article is right that perf per dollar is better, but that’s only because Intel is not making money out of these. As a user that’s a red flag, because if that continues these will be discontinued, and then any investment I do right now in supporting these in my SW stack goes to waste.
Tax on fuel reflects the externalities to the whole of society, and arguably it is not high enough.
So exempting tax (equivalently not charging for externalities) is indeed a subsidy, a transfer of wealth from government (and so society) to the airline operators.
70% of German trains that arrive at their destination are “delayed” (30 min or more).
A train that never arrives is not counted as delayed. About 30% of German trains do not arrive at their destination.
In Switzerland, the stats are single low percentage digits.
After multiple ultimatums, Switzerland stopped allowing German trains into Switzerland. The most popular connection between Switzerland and Germany (Zurich-Munich) has been kept but is now operated by Switzerland even though most of the ride is in Germany.
This is shameful and a lose-lose for all parties, but necessary.
I doubt your stats, where'd you get them from? And there are multiple connections per day from Germany to central Switzerland. Take the Hamburg-Interlaken ICE for example.
