AMD FX-8320E CPU Review: The Other 95W Vishera
by Ian Cutress on January 13, 2015 10:00 AM ESTAMD FX-8320E Conclusion
Anyone building a new performance system today is not exactly spoiled for choice. On the super extreme end, native octo-core processors with threading are in the market. AMD’s high end FX models start by comparison to the Core i5 line but boasts double the threads, albeit at almost double the power consumption. Without looking at the reasons for AMD’s E-series launch, it can be easy to scoff and write these processors off.
AMD’s aim with the FX-8370, FX-8370E and FX-8320E was in part to generate new customers at a lower price portfolio and for those wanting to upgrade their AMD CPU from something lower down the stack. Users who purchased an FX-4100 or FX-6300 can happy get more performance by swapping in a CPU, rather than replacing the motherboard and potentially the DRAM to fit. The 95W power point for the E series allows 8-threads for almost all AM3+ motherboards ever made.
That being said, performance of the CPU is still reminiscent of 2012 when the first FX-8000 CPUs were launched and the 32nm process node on which it is formed. Back in our review of the FX-8370E, our conclusion was that it performed much like the FX-8150 except in power consumption, and it was perhaps a foregone conclusion that the FX-8320E was not going to be any better given the statistics on paper.
A plus point worth noting is that the turbo mode seemed a lot more aggressive than our FX-8150 and FX-8370E numbers. This allowed for higher frequencies in some of our variable threaded benchmarks, and gave some better results which is rather odd. However, it was only for a few select benchmarks in the end, with the pure single threaded ones and the fully multithreaded results still grasping at the FX-8150 numbers. The plus side in all this is that our FX-8150 power consumption measured significantly higher than the TDP (156W vs 125W TDP) and the FX-8320E gave a lower result (86W vs 95W TDP), essentially handing any performance/watt trophy to the FX-8320E.
The big number from our review however was the overclocking potential of our sample. Our FX-8320E overclocked like a beast (yes AMD, you can quote me on that), although your mileage may vary (as long as you quote this bit too). Our single data point gave 4.8 GHz at 1.550 volts for a 50% raise in both frequency and POV-Ray results, although the power consumption was over 3x the stock value (262W vs 86W). Most AMD users would also state that 1.550 volts was a bit high, so at 1.375 volts the CPU still gave 4.5 GHz which is still a good end.
The end result of the FX-8320E will be similar to that of the FX-8370E. Users will have to look at their use case and decide if an iterative CPU upgrade, from one of the earlier FX models or the Phenom II range, makes sense in their future computing. Putting down $150 on a CPU is reasonable enough if you have everything around it, although one might argue that if we add in the MSI 970 Gaming ($100), some memory ($50), some storage ($100) and a mid-range GPU ($150), then gaming at around $650 when you factor in the case and PSU is more than possible, especially when a similarly performing Intel system might cost more. AMD's key metric in this instance is pricing.
The end here will be the same as the FX-8370E review:
At the end of the day, most users feel that AMD needs to upgrade the architecture (and the chipset) to potentially increase performance or reduce power. At some point the architectures of the FX and APU line either need to diverge their separate ways, or there needs to be a hard earned reconciliation attempt to find a node and a manufacturing process suitable for both low power graphics cores and high frequency processor cores. We know about AMD's plans for 2016, dealing with ARM and x86, and the announcements on K12 so far point to AMD targeting servers, embedded markets and ultra-low power client devices. Here's hoping desktop side gets a good boost as well.
Facebook
Twitter
RSS
92 Comments
View All Comments
OrphanageExplosion - Wednesday, January 14, 2015 - link
I played Crysis 3 on an i3-4130 and an i5-4690K. On the jungle stages, there's a night and day difference. I suspect you would notice the same going from a 3570K down to an FX-6300. Most titles are GPU-bound, but CPU can cause frame-rate drops too.Cryio - Wednesday, January 14, 2015 - link
For Offfice an AMD APU A6 is more than enough.aphroken - Wednesday, January 14, 2015 - link
for office, a calculator and typewriter sufficesbarleyguy - Thursday, January 15, 2015 - link
Lame comment, seriously.My work laptop is a high end i7 (Dell 4800 mobile workstation), and my home office machine is an AMD FX-6300. The AMD machine feels every bit as fast.
For typical office applications, there will be no noticeable difference between an AMD and Intel CPU. The storage is generally the bottleneck, so a fast SSD is more worthwhile upgrade than a faster processor.
Even for development (which is why I have a mobile workstation), it's a barely noticeable difference. The longest part of a build is the unit tests, and they are mostly network bound.
AnandTechLies - Wednesday, January 21, 2015 - link
some guy trying to tell people to waste money on an i7. if you buy an i7 it takes up most of the money (unless you have a shit ton of money) and the rest of the Desktop Suffers and AMD processors are actually GOOD. This IMFORMATIVE website lies and has been caught faking their review benchmarks to make people buy intel over AMD, read my username and LISTEN also do not trust Tomshardware it makes up bullshit toD. Lister - Friday, January 23, 2015 - link
Do you think the regular aluminium foil is as good as a tin foil for making hats? Or can the invisible martian invaders read our thoughts through it?happycamperjack - Tuesday, January 13, 2015 - link
I agree! Complained about this before. These gaming benchmarks are some of the worst games to use for both CPU and GPU beside maybe Tomb Raider. Consumer enthusiast CPU market is pretty much dominated by gamers. If you are running irrelevant benchmarks, what's the point then?Suggestion for games to benchmarks:
Dragon Age Inquision: Incredibly well threaded and scaled. Hyperthreading actually gives a big boost in this game. A good indicator benchmark for upcoming Frostbite 3 engine games.
Crysis 3 or Ryse: Son of Rome: Very well threaded and scaled games. Good benchmarks for upcoming crytek games.
Maybes:
Far Cry 3: I'd suggest Far Cry 4 when it's fixed. Another very well threaded game that's a well representation of open world game.
Metal Gear Ground Zeroes: Good benchmark for upcoming FOX engine games such as the popular Pro Evolution series.
I would also suggest AC Unity if it's not so broken for multi GPUs. Yea SLI and Crossfire is still broken this moment.
MapRef41N93W - Wednesday, January 14, 2015 - link
Tomb Raider is horrible for CPU comparisons. One of the absolute worse GPU intensive games for testing CPU on the market. A Pentium G3258 can almost match a 4770k in that game.OrphanageExplosion - Wednesday, January 14, 2015 - link
Tomb Raider is useful for GPU testing as it's very heavy on GPU ramped up to the max but you're right in that its CPU usage is minimal.OrphanageExplosion - Wednesday, January 14, 2015 - link
I actually played Tomb Raider on a G3258 and a 4790K. With a 4.5GHz OC in place (and v-sync active), the Pentium was a total match for the i7. However, at stock speeds, the Pentium saw clear lag in CPU-heavy area - physics etc.However, the benchmarking sequence doesn't have any physics-heavy elements.
Reviewers need to start actually playing games and finding the bottleneck areas rather than just running canned benchmarks that prove very little. If that means fewer datapoints so be it. I'd rather have a smaller amount of useful data rather than a larger amount of meaningless data.