Intel Core i5-4670T

The Intel Core i5-4670T is a 45-watt, four-core Haswell desktop processor released in mid-2013, built on Intel’s 22 nm process with 1,400 million transistors on a 177 mm² die. Its benchmark profile places it near the median of the database, sitting at the 35th percentile of all CPUs, with an average benchmark score of 1266. The data shows a processor that trades raw throughput for efficiency, with its most notable characteristic being a narrow but consistent gap against a cluster of rivals that all land within roughly one percentage point of its average score.

Benchmark Performance

The Core i5-4670T’s average benchmark score of 1266 positions it in a tight cluster of competing parts. It trails the Intel Core i5-6400 by a marginal 0.4% (1271 vs 1266), while leading the AMD Athlon 200GE, Intel Core i3-7100T, and Intel Core i7-3615QM by 0.4%, 0.5%, and 0.5%, respectively. These deltas are effectively negligible in real-world terms, indicating that the 4670T delivers performance statistically indistinguishable from these four rivals in aggregate workloads.

Delving into specific workloads, the Cinebench results reveal a consistent pattern. In Cinebench R23, the 4670T scores 3796 in multi-core and 535 in single-core. The multi-core figure is 7.1 times the single-core score, reflecting the expected scaling from four physical cores without hyper-threading. The Cinebench R20 results (1594 multi-core, 224 single-core) and R15 results (382 multi-core, 53 single-core) follow the same proportional relationship. Geekbench scores of 2563 multi-core and 977 single-core show a 2.6x ratio, which is lower than the Cinebench ratios because Geekbench’s workload mix is less parallelizable.

The 35th percentile ranking is the most important context here. This CPU sits in the lower third of all processors in the database, meaning it is outclassed by the majority of modern parts. However, its proximity to the i5-6400 — a newer Skylake part — suggests that the 4670T’s Haswell architecture remains competitive within its power class. The 45-watt TDP is the defining constraint; the data shows that this processor’s performance is deliberately capped to fit that envelope, sacrificing raw throughput for thermal efficiency.

Who Should Consider It

The benchmark data points toward a narrow set of use cases. For gaming, the single-core Cinebench R23 score of 535 is modest by modern standards, and the 35th percentile overall ranking suggests that CPU-bound titles will likely bottleneck on this processor. Older games that rely on two to four threads may run adequately, but the performance headroom is thin. The integrated Intel HD 4600 graphics further limits gaming potential, as the data does not indicate any discrete GPU pairing in the benchmarks.

For office and productivity workloads, the 4670T is a more reasonable fit. The multi-core scores — 3796 in Cinebench R23 and 2563 in Geekbench — are sufficient for document editing, spreadsheet analysis, and web browsing with multiple tabs. The 6 MB shared L3 cache and dual-channel DDR3 memory support are adequate for these tasks, which are not particularly cache- or bandwidth-sensitive. The 45-watt TDP makes it suitable for compact desktop builds where heat dissipation is a concern.

For content creation, the data is less favorable. Video editing, 3D rendering, and large-scale compilation would stress the four threads, and the Cinebench multi-core scores are roughly one-third of what modern high-end processors achieve. The 4670T is not a productive tool for these workloads, as the benchmark results indicate that it will spend significantly longer on render tasks compared to parts in the 90th percentile. It is best suited for secondary machines, home theater PCs, or basic office systems where the efficiency advantage outweighs the performance deficit.

How It Compares

Intel Core i5-6400: The 4670T trails the i5-6400 by just 0.4% in average score (1266 vs 1271). This is a remarkably small margin given that the i5-6400 is a newer architecture (Skylake) with a higher base clock. The data suggests that the 4670T’s lower clock speeds are partially offset by its Haswell efficiency, but the i5-6400 holds a slight edge that would become more pronounced in single-threaded tasks where its newer core design excels.

AMD Athlon 200GE: The 4670T leads the Athlon 200GE by 0.4% (1266 vs 1261). This is a narrow victory for the Intel part, which has two additional cores (4 vs 2) and a larger cache. The Athlon 200GE is a budget dual-core with simultaneous multithreading, yet it nearly matches the 4670T’s aggregate performance. This indicates that the 4670T’s older architecture and low clock speeds significantly diminish its core-count advantage.

Intel Core i3-7100T: The 4670T edges out the i3-7100T by 0.5% (1266 vs 1260). The i3-7100T is a dual-core part with hyper-threading, giving it four threads, same as the 4670T. The 4670T’s two additional physical cores provide a slight multi-core advantage, but the i3-7100T’s higher IPC (instructions per clock) from the newer Kaby Lake architecture nearly closes the gap. In single-threaded workloads, the i3-7100T would likely pull ahead, despite the aggregate score favoring the 4670T.

Intel Core i7-3615QM: The 4670T leads the i7-3615QM by 0.5% (1266 vs 1260). This is notable because the i7-3615QM is a quad-core mobile processor with hyper-threading, giving it eight threads. Despite having half the thread count, the 4670T’s higher clock speeds (2.30 GHz base, 3.30 GHz boost) compensate, allowing it to match and slightly exceed the older quad-core. In heavily threaded workloads, the i7-3615QM’s extra threads would close the gap, but the aggregate benchmark favors the 4670T.

FAQ

Q: How does the Core i5-4670T compare to the Intel Core i5-6400?

A: The 4670T trails the i5-6400 by a marginal 0.4% in average benchmark score (1266 vs 1271). The two processors are effectively tied in aggregate performance, despite the i5-6400 being a newer architecture.

Q: What is the processor’s single-core performance?

A: In Cinebench R23, the single-core score is 535. In Cinebench R20, it scores 224, and in Cinebench R15, it scores 53. Geekbench single-core is 977.

Q: Is the Core i5-4670T good for multi-threaded workloads?

A: The multi-core scores are 3796 in Cinebench R23, 1594 in R20, and 382 in R15. With only 4 threads, it scales modestly, but the 35th percentile overall ranking indicates it is outclassed by most modern processors in heavily threaded tasks.

Q: What integrated graphics does this processor include?

A: The processor integrates Intel HD 4600 graphics. The benchmark data does not include any GPU-specific tests, so performance conclusions are limited to the CPU scores.

Q: What is the memory and socket support?

A: It supports dual-channel DDR3 memory, uses the Intel Socket 1150, and has PCIe Gen 3. ECC memory is not supported.

Q: When was this processor released, and what is its power draw?

A: The release date is June 1, 2013. The thermal design power (TDP) is 45 watts, which is a defining characteristic of this low-power Haswell part.

Single-Thread vs Multi-Thread Behavior

The split between single-thread and multi-thread performance is the clearest signal in this processor’s profile. In Cinebench R23, the single-core score of 535 versus the multi-core score of 3796 yields a scaling factor of 7.1x across four cores, which is close to the theoretical 8x if all four cores scaled perfectly. This indicates that the Haswell architecture on this chip is well-balanced for parallel workloads that can utilize all four threads, with minimal overhead from synchronization or memory contention.

However, the absolute scores tell a different story. The single-core score of 535 in R23 is low compared to processors in the 90th percentile, which typically exceed 1500. This means that the 4670T will feel sluggish in single-threaded tasks like web browsing with heavy JavaScript, spreadsheet recalculation, or older games that rely on a single primary thread. The 2.30 GHz base clock and 3.30 GHz boost clock are the limiting factors here; the architecture has decent IPC, but the low clocks cap its ceiling.

The Geekbench results show a 2.6x scaling factor (2563 multi-core vs 977 single-core), which is lower than Cinebench because Geekbench’s workload mix includes more memory-bound and latency-sensitive tests. This suggests that the 4670T’s dual-channel DDR3 memory bus and 6 MB shared L3 cache are adequate for the four threads, but they do not provide the bandwidth headroom that newer platforms offer. In practice, this means the processor performs better in sustained compute tasks (like rendering) than in bursty, latency-sensitive workloads (like application launches or database queries).

The practical takeaway from the data is that the 4670T is a balanced quad-core for its era, but the 35th percentile ranking reflects that both its single-thread and multi-thread performance have been surpassed by newer parts. The narrow deltas against its nearest rivals — all within 0.5% — underscore that this processor competes in a very tight performance band, where architectural differences and clock speeds nearly cancel out. For users considering this CPU, the data suggests that its 45-watt efficiency is the primary selling point, not its raw speed, as the benchmark results place it squarely in the lower-midrange of the performance spectrum.