What the November Bottleneck Index Q4 update will measure
The Bottleneck Index is a simple idea with a demanding method: take common CPU and GPU pairings, test or estimate each at 1080p, 1440p and 4K, and summarise which CPUs hold back which GPUs, and by roughly how much, at each resolution. The quarterly update repeats it with the same pairings and method, so changes between quarters reflect hardware and games, not the measuring. That is the point of an index: a single measurement is a snapshot, but a repeated measurement is a trend.
Q4 will measure the same thing Q1 measured: the balance of the standard pairing set across the three resolutions, with results as ranges, not single exact percentages. It will also extend the pairing set to cover hardware that arrived during the year — new tiers and processors that did not exist at Q1. It will not rewrite the method to flatter a new product, or cherry-pick pairings for a launch. An index that changes its method every quarter is not an index; it is a press release with a chart.
This article explains the Q4 update: what it will measure, how the Index is built, and what has changed since Q1. No Q4 numbers are published here. There are no tables of Q4 scores, no ranked lists and no “biggest mover” claims — the data release is a separate publication with its own method note and dataset. If you came for the numbers, this article tells you how to read them. If another page quotes Q4 numbers before the release, treat it with the suspicion any number without a method deserves.
For background, start with our explainer of the Bottleneck Index and the Q1 edition explained, which describes the first release and its limits. The Q2 update and the Q3 update cover the intervening quarters, so Q4 is the fourth measurement in a consistent series, not a fresh start.
How the Index is built
An index is only as trustworthy as its method. The Index starts with a fixed pairing set of CPUs and GPUs builders actually own or ask about, spanning budget, mid-range and high-end parts. Each pairing is assessed at 1080p, 1440p and 4K, because balance changes with resolution — a pairing CPU-limited at 1080p can be GPU-limited at 4K, and one number per pairing would hide the fact builders need most.
For each pairing and resolution, the result is a range, not a single exact percentage. The range reflects honest uncertainty: different games stress the CPU and GPU differently, settings shift the balance, and test systems produce different numbers. A single decimal percentage implies a precision no bottleneck measurement can honestly claim. Our methodology page, how we calculate bottlenecks, explains the same principle for the calculator, and how we built our dataset describes the sources and calibration behind the numbers.
The Index separates what it measures from what it does not. It measures CPU-GPU balance in gaming at three resolutions — not streaming, video editing, 3D rendering, or the quality of a build: a badly cooled PC, a single stick of RAM, or a nearly full hard drive can all make a “balanced” pairing perform badly. Those limits are part of the method; ignoring them is a misreading. Q4 will carry the same limits, stated in the same place, so Q1 and Q4 compare like with like.
In short, the Index measures one thing and leaves the rest to other guides:
| What the Index covers | Measured, or not |
|---|---|
| CPU-GPU balance in gaming at 1080p, 1440p and 4K | Measured, as ranges |
| Streaming and recording loads | Not measured — see the OBS/NVENC guide |
| Video editing and 3D rendering | Not measured — a different workload |
| Cooling, RAM and storage effects | Not measured — they change results |
What changed since Q1
Q4’s hardware context differs from Q1’s, and that is why the update exists. During the year, new graphics-card tiers and processors arrived, and the newest games raised the demands on both — particularly on video memory and CPU performance in crowded scenes. None of that changes the method, but it changes which pairings matter: unusual Q1 combinations may be common by Q4, and balanced Q1 pairings may be stretched by newer games.
Most important is the graphics side: the current generation has settled in, and builders pair its cards with CPUs from several generations. That is the situation the Index describes: the same card with older, current and high-end CPUs, at three resolutions. Q4 will cover those combinations on the same terms as Q1, so the change reflects the hardware, not the measuring. Our RTX 50 series deep dive covers which CPUs can feed the current generation, and the CES 2027 previews of RTX 50 Super, Zen 6 and Nova Lake describe the hardware on the horizon as previews, not measured results.
The game context has changed too. Autumn titles stress the CPU in ways summer titles do not — bigger crowds, larger simulations, more characters tracked at once — and an index measured only on light esports titles would miss that. That is why the Index covers more than one kind of game, and the Q4 game set matters as much as its pairing set. A pairing’s balance is a property of the parts, resolution, settings and game — and any single number hiding those four factors hides the answer.
Limitation: this article explains the Q4 update's method and context. It publishes no Q4 index numbers, no Q4 rankings and no measured results, and nothing in it should be quoted as a Q4 finding. When the Q4 data release is published, it will carry its own method note and dataset — judge the numbers there, not here.
How to use an index update when it lands
When the Q4 data release is published, read it as the method asks. First, find your pairing at your resolution, and read the range. A range spanning “roughly balanced” to “mildly CPU-limited” tells you something real: the answer depends on the game and settings, and your experience may sit anywhere inside it. Second, compare the Q4 range with the Q1 range. If the range moved, ask what moved it — a new game, a driver change, a new card tier — before concluding that your PC got slower.
Use the Index as a starting point, not a verdict. If the Index says your pairing is CPU-limited at 1080p, confirm it with an overlay: CPU near maximum, GPU waiting, frame rate capped below what the card could deliver. If the overlay disagrees, trust it — it measures your PC, your games and your settings, while the Index measures a standardised set that can only approximate them. Our guide to the five-step self-test for CPU bottlenecks shows how to do that check in ten minutes, and what a bottleneck percentage means explains why a range is honest.
Do not use the Index to choose a single “best” pairing. The Index describes balance, not value, availability, power draw, noise or fit — and a perfectly balanced pairing at twice your budget is not a recommendation. Use the Index to narrow the field to balanced pairings at your resolution, then choose by price, power and fit. Run the shortlist through the PC bottleneck calculator, check the power supply and case, and let the Index tell you where the limit sits, so you can spend on the part that removes it.
Frequently asked questions
Where are the Q4 index numbers in this article?
There are none, by design. This article explains what the Q4 update will measure, how the Index is built and what has changed since Q1. The Q4 numbers will be published as a separate data release with their own method note and dataset.
How is the Bottleneck Index built?
From a fixed set of real CPU-GPU pairings, assessed at 1080p, 1440p and 4K, with results as ranges rather than single exact percentages. The pairing set, resolutions and method stay consistent between quarters, so changes reflect hardware and games, not the measuring.
Why does the Index report ranges instead of one percentage?
Because balance depends on the game, settings, resolution and test system — a single decimal percentage implies a precision no bottleneck measurement can honestly claim. A range tells you where the answer sits and how much it varies.
What changed between Q1 and Q4?
The hardware and game context: new graphics-card tiers and processors arrived during the year, and newer games raised demands on video memory and CPU performance in crowded scenes. The method did not change — that is what makes Q4 comparable with Q1.
How should I use the Q4 update when it is published?
Find your pairing at your resolution, read the range, compare it with Q1, then confirm it on your own PC with an overlay check. Use the Index to narrow the field to balanced pairings, then choose by price, power and fit — and check the shortlist in the calculator.