Length and complexity of supply chain as a risk factor for HPC and storage

We’ve seen issues in the past, with massive flooding in Thailand, wreaking havoc on critical components in supply chains. The subsequent demonstration of the basic economics laws of supply and demand did not make users or vendors very happy.

This arose due to a significant over-allocation of one small geographical region to a critical component in offerings. To a degree, this also pushed companies to start looking at how to make this “Somebody Elses Problem” (e.g. move to The Cloud™ and let them sort it out).

There have been other ebbs and flows in this space. There’s been shortages and then overcapacity of memory, NAND, etc. as forecasts mispredict demand. Manufacturing capacity coming online has a latency to it, often measured in multiple year increments.

Couple this to placing manufacturing in the most economically favorable locations, that is, the ones with the least cost associated with such sites, and you now have very (physically) long supply chains. This isn’t a problem when things run well.

But they are not running well.

When supply chains cross national boundaries, the risk cost equation has to include tariffs, duties, slow walking of customs service, etc. Which makes supply chains vulnerable to political winds.

This has downstream impacts upon product offerings, as you can’t build systems to gain revenue, without parts you can’t get. Or if your cost models are blown up as parts costs increase dramatically. Which directly impacts your bottom line, and if you increase prices to offset your costs, your top line.

All of this. From global supply chains. To take advantage of lower cost geos.

The fundamental question is, is the risk to the supply chain, to the bottom and top line, actually worth the cost difference? Put another way … if the supply in question were to go to zero for an unknown extended period of time, larger than a single fiscal year, what would this do to your product, revenue, profitability?

I don’t have the answers to this, though I can note that companies are starting to reconsider the value of lower cost geos, as the material risks to their businesses pile on.

For HPC and storage, most of the manufacturing is done in China. This is, again, due to cost reasons. I suspect, though we wouldn’t see this move quickly, that this may be in the process of changing. I don’t see this happening for nationalism reasons, but for simple cost and risk economics.

I suspect that we may start seeing more fabs and electronics factories built in the US, Canada, and Mexico. The risk in those border crossings are lower, but there as still risks. One needs to balance the risks of political brinksmanship against the ability to deliver product and make profit.