The large api/token providers, and large consumers are all investing in their own hardware. So, they are in an interesting position where the market is growing, and NVIDIA is taking the lion's share of enterprise, but is shrinking at the hyperscaler side (google is a good example as they shift more and more compute to TPU). So, they have a shrinking market share, but its not super visible.

> The large api/token providers, and large consumers are all investing in their own hardware.

Which is absolutely the right move when your latest datacenter's power bill is literally measured in gigawatts. Power-efficient training/inference hardware simply does not look like a GPU at a hardware design level (though admittedly, it looks even less like an ordinary CPU), it's more like something that should run dog slow wrt. max design frequency but then more than make up for that with extreme throughput per watt/low energy expense per elementary operation.

The whole sector of "neuromorphic" hardware design has long shown the broad feasibility of this (and TPUs are already a partial step in that direction), so it looks like this should be an obvious response to current trends in power and cooling demands for big AI workloads.

So the answer is yes, but not to a noticeable amount. Don't worry about protecting your battery life, and charge your phone as needed.

My anecdata fully supports that conclusion.

I tried very hard to keep my battery healthy by only charging my phone to 80%, which resulted in me running out of battery quite often throughout the day, interrupted charging cycles etc., which apparently all were worse than just charging to full (via "optimized charging" which finishes the charge in the later hours) overnight.

I lost about 10% of reported battery health that way in less than a year, switched to just charging overnight or whenever getting close to empty, and have lost all of 3% over the second year of using this phone.

On top of that, swapping an iPhone battery is annoying, but having to worry about all of those battery health concerns is much more tedious.

My understanding is that battery degradation is not linear with time.

So you're still 7% ahead of charging only to 80%.

Nvme pricing is pretty volatile in the past 2 years I’ve seen it move between 2-3x from its low post Covid.

I don’t think the prices have adjusted because of that. Additional during Covid the prices were very high and this is baked into the pricing.

Realistically groq is a great solution but has near impossible requirements for deployment. Just look at how many adapters you need to meet the memory requirements of a small llm. SRAM is fast but small.

I would guess their interconnect technology is what NVIDIA wants. You need something like 75 adapters for an 8b parameter model they had some really interesting tech to make the accelerator to accelerator communication work and scale. They were able to do that well before nvl 72 and they scale to hundreds of adapters since large models require more adapters still.

We will know in a few months.

I would think this is for rental fleets or bike share. The weight and design would seem to make sense for that. Though the single speed seems like and odd choice for that.

This is not true. Almost all firmware is signed by every vendor, and there are standards from Intel and amd on implementation of code signing.

Look up Intel pfr.

Signed ≠ enforced.

At least for 4677 Intel stuff, gigabyte & HP and others let you modify the firmware and flash it.

HPE at least makes you flip a DIP switch, otherwise it complains loudly and halts.

The one vendor mentioned in the comments, AMI, is switching this code base to openbmc. Also it should be noted that often this software is system specific.

The issues were durability, fire rate, and well power.

I don’t know that the first two have changed significantly.

I think that the private carriers are more likely to be helped by this, since they will manage the paperwork.

It’s more likely a set of products that were shipping directly from factories disappears from the market. For example, the direct from factory Halloween costume.

It could end up being a step backwards in living standards and access to daily luxuries.

Old man in a different country: we used to make Hallows'een costumes out of old shirts worn backwards and sacks and stuff. Yes, I'm going back half a century plus but it was fun and involved time with parents.

Best of luck.

Gradual damage is consistent with over heating. I've seen racks of servers do the same thing.

Overall, there is a continued challenge with CPU temperatures that requires much tighter tolerances both in the thermal solution. The torque specs need to be followed and verified that they were met correctly in manufacturing.