$IONQ's decoder claim is in. Tuesday's bar is not.
Graham McCain · Alphaforge Research · Published Sep 23, 2026 · 2:11 pm ET · Revision 1 Scored over 5 trading sessions.
AI summary
IonQ announced on September 22 that it tested a quantum error decoder on a single CPU, adding as little as 0.02% to execution time in circuits with up to 408 logical qubits. The note views this as a technical claim, not revenue, a contract, or guidance. The last session in the library was Monday's close at $40.51, which is 52.1% under the 52-week high.
The historical 5-session excess range for Monday's state is -5.6 / -0.0 / +5.5. Tuesday's session is missing, so the note cannot assess the decoder's impact, with the strongest reading being that the claim does not create a new measured state without updated data. The next checkpoint is the first IONQ session that lands in the library after Monday.
Two paragraphs generated from the note. The authored note below is the record. Numbers only, no side.
1. The state
As of the 2026-09-21 close
Monday's close, the last IONQ session in the library, was $40.51, 52.1% under the 52-week high. Volume was 0.75 times the 20-day average. Tuesday's bar is not in the library.
2. What's new
On September 22 IonQ announced what it called the industry's first end-to-end real-time quantum error decoder on a single off-the-shelf CPU.
3. What that does to the state
Monday's 5-session excess band is -5.6 / -0.0 / +5.5.
300 similar days · 204 stocks · 224 sessions · 1h band
Vs the market, in points
Low
Middle
High
Share up
1 session
-3.0
+0.1
+2.4
50%
5 sessions
-5.6
-0.0
+5.5
50%
10 sessions
-8.9
-0.1
+9.9
49%
The band is the prior close's neighborhood — what followed similar states as of the cutoff.
4. Same state, or a new one
A Tuesday or Wednesday close in the library, outside Monday's 5-session band, would end this abstention.
The next checkpoint is the first IONQ session that lands in the library after Monday. Session counts for the September 23 note start when that bar exists.
IonQ said on September 22 that it had tested an end-to-end quantum error decoder running on one ordinary CPU. The company described circuits of up to 408 logical qubits and more than 31.5 million operations, and said the decoder added as little as 0.02% to execution time under its test noise. That is a technical claim. It is not revenue, a contract, or guidance.
The library does not have Tuesday's IONQ bar. The last completed session on file is Monday, September 21, a close of $40.51. That close is 52.1% under the 52-week high. Monday's volume was 0.75 times the 20-day average. Wednesday's rally, reported in the financial press after the release, is an open session here. It is not a result this note can score.
Because Tuesday is missing, the analog band in this note is Monday's, not the tape people are trading today. Five-session excess on that Monday state is -5.6 / -0.0 / +5.5. Using it as if it described Tuesday would be a false precision. The next honest test is the first session that actually lands in the library, set next to this claim.
Other reading, tape, and sources
The other reading
If Tuesday's bar, once it arrives, is a large move on the decoder release, Monday's quiet close stops being the needle. This note will be wrong to keep waiting only if that bar is already knowable and was missed. It was not in the library at publication.
What would change this
A Tuesday or Wednesday close in the library, outside Monday's 5-session band, would end this abstention. A later filing that puts dollars on the decoder would make the press release an operating event.
The tape
Monday's close, the last IONQ session in the library, was $40.51, 52.1% under the 52-week high. Volume was 0.75 times the 20-day average. Tuesday's bar is not in the library.
What was filed
On September 22 IonQ announced what it called the industry's first end-to-end real-time quantum error decoder on a single off-the-shelf CPU. The release cites a test of up to 408 logical qubits and more than 31.5 million operations, with as little as 0.02% added time. No revenue figure is in the release. Wednesday's price move is not a completed session in this note.
The numbers around the move
Monday's 5-session excess band is -5.6 / -0.0 / +5.5. It is not a band for Tuesday, because Tuesday is absent. The decoder claim does not by itself create a new measured state.
What the stock did
No completed session after the release is in the library, so this note cannot say the tape left or stayed in Monday's neighborhood. That sentence waits for the next bar we actually have.
The comparison
Five-session excess -5.6 / -0.0 / +5.5 across 300 analogs on the Monday, September 21 state. One session -3.0 / +0.1 / +2.4. Ten sessions -8.9 / -0.1 / +9.9. These ranges belong to Monday. They are not a forecast and imply no side.
Sources
Monday tape and band from the September 21 state packet. September 22 has no IONQ daily bar in the library. The decoder claim is from IonQ's own release. Wednesday's intraday price is deliberately not scored.
