IPC (instructions per clock): why a 4.5 GHz Zen 5 beats a 5 GHz Zen 2 and how IPC grew per generation
IPC (instructions per clock) measures how much work a CPU core completes per cycle; single-thread performance is IPC × clock speed, and for games it is the primary determinant of frame-preparation speed. Every core generation raises IPC by 10–20 % through wider decode, better branch prediction and larger caches, which is why a newer CPU at lower GHz outperforms an older one at higher GHz.
instructions per clock · single-thread performance · per-core performance
Key figures
From the FPSHQ database, as of Oct 13, 2026
What raises IPC
Wider front-ends (decode 6–8 instructions per cycle), deeper out-of-order windows, more execution ports, smarter branch predictors, bigger L1/L2 caches and faster memory access all let the core finish more instructions per cycle. IPC is workload-dependent — a game's branchy, cache-missing code gets different gains than a vectorised encoder — so vendors quote an average across benchmarks; for games, cache and memory latency carry a larger share.
Generational IPC gains (vendor figures)
Combined with clock changes, these produce the single-thread index in the FPSHQ CPU catalog; the gaming index differs because cache and latency weigh more there.
| Core | Year | IPC gain vs previous |
|---|---|---|
| AMD Zen 2 | 2019 | +15 % vs Zen+ |
| AMD Zen 3 | 2020 | +19 % |
| AMD Zen 4 | 2022 | +13 % |
| AMD Zen 5 | 2024 | +16 % |
| Intel Golden Cove (12th gen) | 2021 | +19 % vs Cypress Cove |
| Intel Raptor Cove (13th/14th) | 2022 | ~0 % (clocks + cache) |
| Intel Lion Cove (Core Ultra 200) | 2024 | +9 % (but higher memory latency) |
Reading CPU specs with IPC in mind
The compare tool on FPSHQ shows single-thread and gaming indices side by side so IPC and clock differences are visible in one place.
- Compare generations by benchmark results, not GHz: a Ryzen 5 7600 at 5.1 GHz beats a Ryzen 7 2700X at 4.3 GHz by ~60 % per core
- Within a generation, clock speed differentiates models (i5 vs i7 vs i9)
- High IPC with poor memory latency (Arrow Lake) underperforms in games relative to synthetic tests
- Laptop chips share IPC with desktop but run lower clocks under power limits
On your PC
Enter your GPU (and CPU) — we calculate the effect of "IPC" with the unified FPSHQ model using Cyberpunk 2077 as the example.
Save your PC in the account · Detect hardware with the app · Unified FPSHQ model estimate, not a measurement.
Where to use this on FPSHQ
Forum discussions
No topics with this term yet — be the first to ask.
What is IPC?
IPC (instructions per clock) measures how much work a CPU core completes per cycle; single-thread performance is IPC × clock speed, and for games it is the primary determinant of frame-preparation speed. Every core generation raises IPC by 10–20 % through wider decode, better branch prediction and larger caches, which is why a newer CPU at lower GHz outperforms an older one at higher GHz.
- most common socket
- LGA1700 · 63
- CPUs with 3D V-Cache
- 12
- median L3 since 2023
- 32 MB
- median L3 in 2015–2017
- 9 MB
FAQ
What is IPC in simple terms?
Is a higher GHz CPU always faster?
How much does IPC improve each generation?
Why does Arrow Lake have higher IPC but lower gaming FPS than 14th gen?
Sources
- AMD Zen architecture (2025)
- Intel Core Ultra 200S (2025)
FPSHQ Team · methodology · methodology · Updated
✏️ Found a mistake? Suggest an edit
Get the FPSHQ app — it detects your hardware and shows FPS in your games.
💻 Windows app