• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Pentium K gaming review from Bit-Tech

Rofflay

Rofflay

Rofflay

Permabanned
Joined
28 Nov 2006
Posts
5,750
Location
N Ireland
http://www.bit-tech.net/hardware/2014/06/24/intel-pentium-g3258-review/5



Pretty nice for just £60 i actually was tempted to do a £600 Prodigy build i priced with a Pentium K and some DDR3 2133 memory and a 270X card. If it went to 4.8 and the memory could be got up it is a screamer at the minute until all the engines sort out multi-threading. I imagine this is a dream for World of Warcraft players as well who tend to sit on ancient computers.


Anyone building a rig with one?
 
Currently putting a hadron air mitx build together


Just waiting on this little sweety launching pre order in place :)
 
Last edited:
thing is second hand i5 for 30 quid more will be a lot better. i wouldn't use one for a system build. for a bit of fun and cheap as poss maybe.

should be 50 quid.
 
I think a lot of us will get them for fun.
OC them till they scream for mercy.

Perhaps 8pack could do a comp.
winner gets a 4790k, cherry picked of course.
 
ubersonic said:
A £60 Intel CPU is beating AMD's top CPU in gaming benchmarks, what parallel dimension have I stumbled into? lmao.
Click to expand...

LMAO,that Bit-tech test Skyrim and Total War and call it a gaming test(two games). No frametimes at all.

Whats the point of all the frametime measurements that were done for graphics cards tests to measure stuttering,when they are not done for CPU tests??

Lets look at some results:

http://www.tomshardware.co.uk/pentium-g3258-overclocking-performance,review-32974.html

Look at where the Core i3 is in the review. Look at where the ancient Athlon II X4 750K is,even with a subpar overclock. People seem to ignore the frametime measurements for some weird reason.

A lot of games are not that single threaded or limited by even older CPUs- its just a myth propagated on forums. If that was the case the G3258 would be destroying the Core i3 4330 when overclocked and HT would make zero difference.

It does not with a Geforce Titan.

E8400 against Q6600 - the former had better IPC,higher base clockspeeds,and generally would overclock better(20% better on average). Yet the Q6600 lasted far longer. Games were less threaded then,than they are now,and it still did not help.

HT is making a big difference,and it seems that Haswell shows better HT scaling on the dual cores than previous generations(I would definitely like to see more reviews with frametimes especially against he FX63** since the improvements might be enough to match it in more modern games).
 
Last edited:
Cheapskates could pair them up with B87/H87 boards.
ASUS released the BIOS updates, others are bound to follow.
 
CAT-THE-FIFTH said:
LMAO,that Bit-tech test Skyrim and Total War and call it a gaming test(two games). No frametimes at all.

Whats the point of all the frametime measurements that were done for graphics cards tests to measure stuttering,when they are not done for CPU tests??

Lets look at some results:

http://www.tomshardware.co.uk/pentium-g3258-overclocking-performance,review-32974.html

Look at where the Core i3 is in the review. Look at where the ancient Athlon II X4 750K is,even with a subpar overclock. People seem to ignore the frametime measurements for some weird reason.

A lot of games are not that single threaded or limited by even older CPUs- its just a myth propagated on forums. If that was the case the G3258 would be destroying the Core i3 4330 when overclocked.

It does not with a Geforce Titan.

E8400 against Q6600 - the former had better IPC,higher base clockspeeds,and generally would overclock better(20% better on average). Yet the Q6600 lasted far longer. Games were less threaded then,than they are now,and it still did not help.

HT is making a big difference,and it seems that Haswell shows better HT scaling on the dual cores than previous generations.
Click to expand...


Why would they measure frametimes on a £60 cpu though? This is a very niche product for a specific criteria which is why i thought thier choice of games was relative. Skyrm, Shogun II and World of Warcraft are ideal games to show off because they are at most single or dual threaded and get good gains from IMC and memory speeds.


Those games also are played by a lot of people on steam, WAY more than the proper multithreaded titles in all honestly. Do you think someone playing Dota 2 will want to spend extra? You can price a full microprodigy system with Z87 Pentium K and 270X for £600 i think i even had a copy of win8 bundled. Even in Bioshock Infinite it will do Ultra 1080p at a full 60fps with DOF off if you can match 4800mhz.
 
Last edited:
Rofflay said:
Why would they measure frametimes on a £60 cpu though? This is a very niche product for a specific criteria which is why i thought thier choice of games was relative. Skyrm, Shogun II and World of Warcraft are ideal games to show off because they are at most single or dual threaded and get good gains from IMC and memory speeds.


Those games also are played by a lot of people on steam, WAY more than the proper multithreaded titles in all honestly. Do you think someone playing Dota 2 will want to spend extra? You can price a full microprodigy system with Z87 Pentium K and 270X for £600 i think i even had a copy of win8 bundled. Even in Bioshock Infinite it will do Ultra 1080p at a full 60fps with DOF off if you can match 4800mhz.
Click to expand...


Frametimes are an indication of stuttering,which is why they need to be included. Plus I don't disagree about some games like WoT and WoW(although WoW does use more than two threads) to a degree. This would be great for them. I see those as lightly threaded games,which do have issues.

However,for DOTA2,I don't agree as much.

A lot of people playing DOTA2 and LoL(like many people I know),can run them fine on old desktops and laptops.

I ran CS:GO,DOTA2 and LOL at 1920X1080 max settings on a blasted ancient HD5850 and a A6-3670K and a Q6600(former is a relative's PC and the latter my ancient Shuttle system) at stock.

An A10 7850K which won't win awards for CPU and GPU speed can run all those games at reasonable settings.

Forget the G3258 - I would get a £30 Celeron instead if those are the kind of games you want to play. That plus an HD7770 would give you great framerates.

There is a reason why massively popular games,are well popular - they can run on pretty crap hardware(by enthusiast standards) and have cartoony art styles which do well even with graphics settings dropped.

Lets look another game which I play a lot,namely Diablo3. It is lightly threaded and yet when framerates do crash during MP rifts with loads of bosses sometimes,it happens for everyone,ie,those with Core i7s,Core i5s,FX,Core2 CPUs etc. Its because the major limitations at that point are on the server side.

Yet in SP testing I did with an anaemic and ancient A6-3670K IGP,I was getting above 30FPS at 1440X900 through entire sections of the game:

http://i559.photobucket.com/albums/ss33/CAT-THE-FIFTH/D3benchmark1.png
http://i559.photobucket.com/albums/ss33/CAT-THE-FIFTH/D3benchmark2.png

Loads of spawned packs,at high settings(no shadows enabled) and the game is hosted online even in SP mode.

Plus,it is not a proper game review if a range of games is not tested. You could test Thief and Watch Dogs then,and make the Pentium dual core look crap(stuttering) and the Core i7 4790K look like the dogs ********.
 
Last edited:
What games would you say need to be tested Cat?
As i'll being doing just that, with a single 290. Overclocked the GPU to try and bottleneck the CPUs as much as i can :)

I've got a few games, just takes a while to download the larger ones :(
Backed a lot of them up, just need to find the drive they're sitting on :D
 
I heard they have limited dram freq to 1333mhz aswell?

its decent chip but needs hyperthreading to be truly awesome imo
 
wazza300 said:
I heard they have limited dram freq to 1333mhz aswell?

its decent chip but needs hyperthreading to be truly awesome imo
Click to expand...

If that's true it seems to be a bit of a drag.
Otherwise it seems like a dream for entry level gaming/steambox type users.
 
You do not need frametimes it will not stutter as long as it hits the minimum 60fps. I have never heard of a cpu that has stutter. I would imagine it would be called defective for gaming and i have owned budget chips before and played games on them from the 2500M to E6300.


Putting it in a benchmark with a 290x or something is just wrong lol. The same can be said of running it in a game that clearly uses above 2-3 cores like Battleturd series. Although for the money i mentioned for possible 4.8 2400mhz DDR3 and a 270X OC at £600 what could compete? You could probably buy a lone 4790k 2400mhz Z97 rig and use the IGP 5000/6000 or whatever they use now but from what i can find they do quite bad at Skrym,wow,Bioshock compared to the £113 270X OC.


I heard they have limited dram freq to 1333mhz aswell?
Click to expand...

The last bit of info i had said otherwise i believe i saw it with 2400mhz memory. The OCUK site said it works up to 2666+ on Z87/97. Bit-Tech state in the test system they used 1866+.
 
Last edited:
Building my daughter a budget mitx gaming rig round one of these

Z87 mobo hoping to push 4.5 ghz minimum paired with maybe a 280x for 1080p gaming
 
Get the 270x OC (£115) and the cheap 500B EVGA psu's and there is £50 sticks of 8GB 2133mhz floating around too. You can build her one that will do Bioshock 1080p max pretty much if you luck out on the overclocks. But your going to come close with a 500w gold. Still you cannot argue for a £36 quid psu.


Add in windows 8 and your gold you really should not be crossing the £650 mark at all. I reckon £570 all in atm.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom