PSU tier list 2.0

Page 176 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
I was thinking of getting an EVGA supernova GS 850W, which is tier one. However, upon reading some of the replies in this thread wondered whether a tier two would be acceptable for getting a 4790k to 5Ghz and and r9 390 to +150 core. I currently have £85 and don't really want to wait much longer before getting a new psu, the EVGA is £99.99 from Amazon. (I currently have an EZCool (Tier 5)

:O
 


I don't care what power supply you use getting a 4790K to 5Ghz will require a golden chip and a custom water loop.
 
anort, I am currently running 4.9ghz @ 1.313v on a be quiet! shadow rock 2 air cooler and am going to get a kraken x61 soon, I do believe I won the silicon lottery grand prize with my chip. But should I go down to tier 2?
 
The tiers are kinda important, but mainly as a source of reference. A tier2 platinum will have better efficiency and maybe even better output ripple suppression than a tier1 gold unit, but is rated tier2 simply because some of the internal parts are not of the highest quality. A good example is the older corsair RM 750. Built by Chicony, it was excellent quality build, amazing outputs on par or better than Seasonic X, yet ranked tier4 because of thermal properties causing shutdowns under maximum load conditions. So there's that to consider.

High OC is all about stability, mainly voltage regulation. You won't get stable OC of 5GHz on a unit with funky v-droop and uneven ripple. For that, a tier1 unit is more purposed, but regardless of tier, the units showing consistent quality ripple suppression take precedence. The G2 is slightly better off there than the GS.

It's a psu, that you will demand top performance from. You will, by necessity, need to pay a premium for that. Don't skimp, its just not worth it.
 


Thanks for the reply, I am aware of voltage ripple playing a big part in over clock stability and I was considering the g2 also (It is easier to buy sleeved cables for). However I was unsure as to whether it was worth an extra £20.
 
Corsair RMx on tests has shown to have tighter ripple than the EVGA G2. Also, It would not be a bad idea to look into something overkill because if it is overkill ripple should be lower at that load level. Efficiency might take a hit, but ripple could be even better.
 
The FSP Hydro X units have been released today. Very high quality.

Hydro X 450 review: http://www.tomshardware.com/reviews/fsp-hydro-x-450-power-supply,4495.html
Hydro X 650 review: https://www.kitguru.net/components/power-supplies/zardon/fsp-hydro-x-650-power-supply-review/

These should not be thought of as "budget" units but high-performing mainstream units. Compared to the Hydro G, they are not modular, but their performance is not at all far behind, as they perform very well. All 100% Japanese electrolytic capacitors and double ball bearing fan that has proved to be a very quiet unit, which is something that is often overlooked. The 450W has an MSRP of $79.99, 550W has an MSRP of $89.99, and 650W $99.99. Since most units end up selling cheaper than their MSRP, these can be competitive units.
 
I have a 550w G2 powering my STRIX 970 at high OC and a 3770k sitting comfortably at 4.6GHz 1.232v. Took it easily to 4.8GHz at 1.28v, so if I could be bothered it'd prolly go higher. This batch has seen more than a few at 5-5.1GHz at less than 1.4v. The G2 barely ever kicks on the fan, and even then it's not heard over the gpu/cpu fans (kraken x61). Would I recommend the extra £20 for a G2 over the GS? In a heartbeat. Not to knock the GS, that's still a great unit, but I'm biased in a big way lol.
 
I agree, the 550 G2 is the best of the G2 series in my opinion. It has tighter ripple and doesn't drop the PWR_OK signal after the voltages go out of spec.

It seems the 550 G2 is about 43db with a power of 550W. Not the quietest but certainly fine.
DnQBc6H.png
 
43 db(A) is unrealistic. That's the psu sitting on a test bench, wide open with a microphone sitting 1m away. Not a psu sitting in an enclosed case where if following any sort of recommended guidelines, the psu will never see close to 100% load, more like 70% and extremely unlikely to see case temps of 45°C to ramp up the fan to max.

At 1m away in my case, my 550w G2 is inaudible at anything close to idle loads and is totally inaudible over the cpu/gpu fans at gaming loads. At idle, the only way I can tell the pc is even on is the damned bright power/hdd led on the top edge of the case
 


But units using Jun Fu capacitors and ChengX like the Cooler Master G650M only get 1 point off with a 9/10 build score with that cheap junk.

Then again, it's not Jonny doing the reviews anymore.
 


...PSU reviews test dishonest product labeling.
 
Yes, they do. For CM to still engage in this practice with its "budget" units, and some of its fans, yields a "Corporate Integrity" score of 0. Unlike all other score components, which are additive, this one is a multiplier. Oops. This is Willful Wrongdoing, for which someone should be in a jail cell or tomb. Until that happens, CM remains on my "Do Not Buy" list. Fortunately, in any market where their products are found, competitors offer similar or better products for similar or lower prices. There's no reason to buy Coolermaster ANYTHING.
 


Not quite nitpicking IMO. He's being illogical. After all, there's a 24 pin and a (4/6 pin? CPU) plug that you MUST USE in any case (no puns intended). Anything else could go unused, so I don't see a reason for .5 deduction.
 
And such features should be for the user to decide, not the reviewer. Quality and performance should be the only things a PSU should be scored on. Not saying the reviewer cannot list what they personally do, or don't, like, as that will provide the reader information, for them to decide if they really care about it or not. It, however, should not be part of the scoring system.
 


Even what is considered "quality" depends much on the end user. For instance, some users who get very frequent brownouts may not want to purchase a unit with PWR_OK troubles, but for someone who virtually never has any issues, the unit is perfectly "quality". Or if someone has a UPS, having low AC_LOSS to PWR_OK time could be trouble if transfer time is longer.

Some people care about noise and others don't. Some people may run a boatload of hard drives which would make the minor rails actually significant, whereas we primarily look at the 12V rail for its performance. Or those people running off configurations of hard drives or 5V PCI cards may have to look into a unit with a particular type of cross-loading performance, whereas another user won't.

Some people may live in hotter areas where things like Japanese caps are more important and units that can tolerate heat. For others with lower ambient temps, whether the unit is rated to 40C or 50C wouldn't matter much (and even then these temperature ratings are very loose and not very standardized, though ATX spec recommends 50C).

Some people who want to do heavy overclocking may think of a "quality" PSU as one with very low ripple, whereas someone who does not overclock may be more concerned about overall reliability in the long run. Or maybe that person who overclocks wants a PSU with a rather high 12V; I don't know if it's rumor or truth, but I've heard PSUs with 12V somewhat higher can be beneficial to heavy duty overclocking.

Let's not forget that gamers would be looking at transient response moreso than an office user or somebody whose PC is at constant 100% load, such as a constant rendering or editing machine. So what really is quality depends just as much on the end user as does functionality.
 
I've been reading a fascinating book.
oo6TZ0K.png

It's very interesting, aside from the author's persistent electricity falsehoods and misconceptions he makes which make me go insane (mostly due to his persistent usage of the word "electricity" rather than talking about charge or energy, which darkens his understanding of things like DC and AC). But the book is very historical, and I'm learning a lot from it so far; I'm on Page 69, about to go onto Nikola Tesla, whose works have been recommended to me before. If anybody is interested in this book, I recommend you give it a shot (and focus on the historic context, please, and not the author's misconceptions of scientific things).

It's a very relative book to power supplies, because if it wasn't for AC being the standard, our computer power supplies would be much different (mostly with less stuff).
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom