Member Info for InverseMe

Very consistent pattern of this stock initially jumping after good news and then steadily falling back to where it was before said news. A lot of patience required with this one but I've been holding for a couple of years already so nothing new haha

This has even more potential than IQE imo. And I just realised a 1200% gain IQE haha

I think the most likely outcome is VPEC buys IQE Taiwan and Lumentum buys the rest.

In the past I could search IQE or IQE Taiwan in Google Maps and it would take me to 高平磊晶科技股份有限公司. Now it will only find the building if I search the Chinese name.

Upon further investigation, I can find the building when entering the address and it looks like the IQE branding has been scrubbed. I'm pretty sure when I searched the building in the past it showed IQE but now there is only the generic Chinese name (Gaoping Leijing).

Interesting thing I noticed: IQE Taiwan is no longer searchable on Google Maps. Could indicate news of a sale is imminent

To be honest, I'm still a bit nervous. If the "Edge AI" ASIC, which we already know had tape-out delayed by the customer, ends up being cancelled then that will leave a big hole in the finances.

Hopefully I'm wrong about the application of this chip, my current thinking is its designed for Oxford Nanopore SmidgION and there are signs this project either has or will be axed.

Surely this is the killer app ONT has been waiting for?

https://deepmind.google/blog/alphagenome-ai-for-better-understanding-the-genome/

Currently trying to buy more but T212 doesn't seem to want to allocate me any shares. Anyone else struggling to get orders filled?

Hi Double-Bubbler, just wanted to let you know that your work on Ensilica is much appreciated. Having thought about the case for a dual US listing, I can only see it happening if the company can get close to a ~$1B valuation and generate some hype as a 'UK tech unicorn'. Not sure how this happens unless they get a big design win somewhere in the hundreds of millions of dollars from a big-name client and need to raise capital for taping out. It's all about momentum in the US markets at the moment so the share price needs to at least 5-10x over the next couple of years before we can even think of a possible US listing imo. On a side note, there's some weirdness going on with the brokerages. For a while you could only buy 1512 shares on T212 but I checked earlier and they increased it to 11150.

Just thought it was worth mentioning that the JLR production halt after the cyber attack may impact the upcoming results. Their biggest supply contract at the moment is $40M over 6 years for an automotive chip and it is thought that JLR is the customer. With some admittedly naive assumptions, this works out to a revenue hit of ~£500k per month if JLR halts orders.

There are definitely areas for growth such as power electronics like you mentioned, but China is leading the way on solar, energy storage, electric motors etc. Just as the USA is looking to cut China out if its supply chains, China is decoupling from the US and companies like ADI and TI still get ~25% of revenue from China. The way Ive decided to play this theme is by owning Aixtron (yet another WFE co. haha) because they claim to have 90% market share for GaN tools and around 25% for SiC. GaN in particular will be critical for future datacentre infrastructure. I've largely given up on 'value' for now, it seems the only way to make money in this market is to ride the current wave of momentum until the AI bubble inevitably pops.

I made a mess of that Nvidia maths haha. Adding $5B revenue per quarter means extra $50B per year, leading to an extra $37.5B net profit per year with 75% margin. This leads to $280B FY30 profit or ~$180B with 50% net margin assuming the same revenue growth. With 30 PE ratio, with 75% margin you get $8.4T valuation so you double your money over the next 5 years compared to $5.4T with 50% margin and only 5% CAGR but PE probably comes down to around 25 in that case and you end up with the same price as you pay today. I know I should be talking about Ensilica lol, just wanted to give everyone a very rough idea of how I approach these kinds of things.

I was so salty about AWE lol, acquired for $2.4 billion while the likes of Credo and Astera labs are currently $30-40 billion market cap. Thing is, recently it became clear that AWE likely wouldn't have survived, their 224G SerDes and UCIe is apparently worse than Synopsys and in that case they don't have much of an IP business. They also seemingly have no hope of competing with Broadcom in ASIC development and the Banias Labs oDSP business is going to be killed by co-packaged optics before it even gets started.

As you said Dartron, the current valuations of most of the US listed companies in this space are pretty rich, pretty much pricing in the next 5 years of growth for the likes of Broadcom and Nvidia. Interestingly, I was looking at Nvidia's financial statements and their revenue growth is surprisingly linear, adding ~$5 billion per quarter for the past 2 years resulting of an expectation of ~$54 billion for their Q3 2026. Naively extrapolating we get ~$200B FY26 revenue. Their net margin is something ridiculous like 75% (the past two quarters were 46% and 56% due to China writedowns). Putting all that together net profit should be about $130B for FY26 which is a 30x PE multiple at the current $4.2T market cap. Keeping the extrapolation going you get $220B FY27 profit, $280B FY28, ... adding $60 billion profit each year so $400B by FY30. Obviously there are questions about how much infrastructure is actually needed, losing access to China (33% of the market), competition from ASICs and so on. Not to mention this assumes insane 75% margins. As soon as the margin drops to 50% with the same assumption for revenue growth, FY30 profit drops 33% to ~$250B meaning 75% increase in share price (15% CAGR) with a 30 PE ratio.

All this to say you can make an argument that Nvidia's current valuation is pretty reasonable.

My portfolio is heavily geared towards WFE. Currently the only US company in my portfolio is Onto Innovation (I'm very confident on this one). I was also going to add Applied Materials but went for Tokyo Electron in the end due to greater NAND exposure and has been on an amazing run lately (Lam, KLA and ASML all at a bit of a premium for my taste and perhaps deservedly so). I'm also holding Soitec (with the idea of it eventually giving high exposure to silicon photonics) but that hasn't worked out too well for me yet and probably won't until closer to the 2028 Nvidia Feynman ramp, if ever lol.

I'm personally not a fan of the IDMs you mentioned, too commoditised for my taste and will probably suffer from oversupply and low margins for the foreseeable future due to Chinese players entering the market. Would be interested to hear your thoughts on this Dartron as I imagine your optimism is founded by your technical background.

Oh I missed a couple more notable acquisitions: Cambridge Silicon Radio was also acquired by Qualcomm and Dialog Semiconductor by Renesas. Not to say such acquisitions are inherently bad for the companies/country, it's likely all these companies are better off under than they would be if they were still publicly listed in the UK and just goes to show the sorry state of equity markets in this country. I would argue the only reason the UK is still somewhat relevant in AI is down to Google's acquisition of Deepmind under the condition that the HQ and operations remain in the UK. ARM is also important but their position in the datacentre is under attack and looks even more precarious after the announcement from Nvidia and Intel yesterday.

This is my last remaining UK holding. The state of the LSE is really sad, basically zero exposure to AI/semiconductors. What little there was has been sold off/gone broke or a combination of the two. ARM sold to Softbank, Imagination Technologies sold to China, Graphcore sold to Softbank, Alphawave sold to Qualcomm, Sondrel (now Aion Silicon) taken private, IQE looking for a seller, and now Ensilica languishing on AIM with no investor backing will also likely be an acquistion target for somebody in the future.

My money is probably better placed elsewhere but I still hold the tiniest bit of hope that a pickup in the auto and industrial markets expected next year and developments in the satellite business will allow for significant growth in the coming years. The trouble is the company is so cash poor that even if they have the technical capability, the scope and scale of projects they can work on is fundamentally limited and they will gradually fall behind better-funded competitors. The LSE is different to the American markets because there is a strange fixation on profits in the immediate term in UK markets at the expense of faster growth in the expectation of much larger longer term profits in the US and they give their companies the capital to have a reasonable chance of executing.

While Ensilica aiming to be cash positive by the end of next year is a good aim, after all we invest in order to earn a share of future cash flows, this comes at the expense of throttling growth and their business is all about scale because scale leads to higher margins. Anyway, I'm just ranting now but hopes this spurs conversation.

I guess people expect another bid, its like a 2p call option

183p is extremely disappointing, surprised ARM walked away if that was the asking price.

Well there we are then, Qualcomm have got themselves a steal.

I believe this extension is also due in part to Qualcomm's Q2 2025 earnings release happening tomorrow (30th April) after the US market closes. Qualcomm is currently in their quiet period and probably wanted to avoid having to disclose anything about the acquisition just before the call.