Intel Pentium 20th Anniversary Edition G3258 CPU Review

CPU by leeghoofd @ 2014-06-16

Intel announced earlier this year the Haswell refresh lineup together with the release of their latest mainstream chipset, the Z97. This refresh release was overshadowed by the eagerly awaited high end K SKU Devils Canyon and the low end 20th Anniversary Pentium processor; especially the latter is a very interesting product as it now sports a fully unlocked multiplier and retails at an insane low price. We have to thank the boys from Denmark Zzolio and Riska which allowed us to spend some time with their G3258 Pentium ES processor.

  • prev
  • next

Pentium Cache Test

Most will wonder how does this Pentium generation perform clock per clock versus more high end offerings. The Pentium G3258 is equipped with 3MB of Level 3 cache. Other mainstream or even the high end Haswell processors sport 6 or even 8MB.  How important is the Level 3 cache ? Is it still as important as in the socket 775 days?

We compare the G3258 versus an under clocked i5-4670K and i7-4770K. On both the K SKUs we have disabled 2 cores and on the i7-4770K we also were required to disable the Hyper Threading feature.

 

 

SuperPi 32M loves bandwidth and a 5 sec gain versus the 3MB cache processor might not seem much , though this is the difference between a top 10 or top 100 spot. Cinebench R10 receives a small boost from 3 to 6MB cache. More L3 cache seems insignificant for this benchmark.

 

 

The OpenGL test benefits highly from the extra cache, we've conducted each test 3 times (after a reboot) and got similar outcome test after test. The Allbenchmark Catzilla 3D test also gets a small boost from the added 3MB cache over the L3 cache of the Pentium G3258, though nowhere as spectacular as with Cinebench R11.5

 

 

 

 

 

The three game test we conducted at the high detail preset show a slight FPS improvement from 3 to 6MB cache. Only the Sleeping Dogs benchmark seems to scale big time with an added 6fps on the High detail preset. Once Ultra mode is selected the GPU bottleneck jumps in.

 

  • prev
  • next