Pacific Buying Challenges: Bifrost & Topaz - Part 1

Right now there is a swarm of buyers seeking Bifrost and Topaz capacity. A former Google subsea cable guy warned me some time ago that these projects would be a big disappointment to the wholesale community.

Google is keeping much of the Topaz spectrum for itself with one major transaction with a government entity. The main providers in the market are MOX and Telstra. I believe each owns a fibre pair with Mox doing a fair amount of spectrum sales in the form of quarter fibre pair sales.

To give you a sense of pricing, a 15 year 100G wave pricing between Tokyo Equinix and Seattle Westin Building is slightly over $3 million upfront with 4% O&M. Ignoring discounting, that's $27K a month! 😂 I estimate most 1 year leases will be in the upper $30K to $50K MRC range. The route is extremely sexy because it is the lowest latency stable route between Tokyo and Seattle, lands in Canada, which Canadians love, and is diverse to other Japan/US cables. Of course, the lowest latency cable is PC-1, but repairs have taken several months per outage, which is catastrophic, particularly for the financial industry. Trading firms move market data and orders between Tokyo and the CME data center in Aurora, Illinois, via PC-1 and Topaz. I think that general bandwidth consumers like ISPs are much better off using TGN-Pacific, which has an excellent uptime history, reasonable latency, and much lower pricing than the Premium Topaz route.

Bifrost is even worse in terms of pricing. It offers one of the first single cable links between Singapore and the US. It is a big deal because normally one must use a Singapore to Japan cable and then use another cable for the Tokyo to US hop. So cost is high and so is latency. But Google and META own Bifrost 100%. It appears that the Dynamic Duo has decided to keep most capacity for their own internal traffic. A few carriers have purchased fibre pairs. Two fibre pairs. So we again end up in a high demand, but low supply situation.

Now Bifrost is not RFS yet. Reliable sources tell me 4Q2025 or more likely 1Q2026. Now one can route traffic from Singapore to Guam to the US today. It is a longer route, but it is available at the 10G level, which will not initially be the case for Bifrost. Existing Singapore/Guam/US pricing suggests Bifrost 100Gs will be in the mid-to-upper 40s. With undoubtedly some transactions exceeding $50K. It is important to emphasize Bifrost's unique advantages. It is the only one cable solution that directly connects Singapore with the US. Moreover, it avoids the South China Sea. China claims this sea in flagrant violation in international law and has created artificial military base islands to enforce it. It's a good reason for avoiding the South China Sea. Moreover, China insists on the right to issue permits for cable deployment in these waters and given its military presence it is not possible to ignore their demands. Permitting takes a long time.