Overclock.net › Forums › Industry News › Hardware News › [TR] Exploring the impact of memory speed on Sandy Bridge performance
New Posts  All Forums:Forum Nav:

[TR] Exploring the impact of memory speed on Sandy Bridge performance - Page 6

post #51 of 77
Quote:
Originally Posted by xxbassplayerxx View Post
There is no 1:1 ratio with 1366, 1156, or 1155.



Still matters on all platforms for Super Pi and the like.



Amount of memory wouldn't do anything for me, but quicker memory makes a huge difference. For a super 32M run, a kit of Elpida Hypers can shave off 10-15 seconds. For Geekbench, it can raise my score ~1000 or more points.

These sticks are 1800 6-6-6 17 1T 32M stable and 2133 7-7-6 19 1T Geekbench stable. I'm sure I can do 32M at 2133 7-7-6 19 as well, but unfortunately my CPU's IMC can't do it as Super Pi 32M is one hell of a load (and it takes a long time).

If Elpida made 3x1GB Hyper kits, I probably would have purchased those. If they did, I never saw them!
Ahh, just noticed you are an HWBOT guru. Ok, that's a legit need for high-speed ram.
post #52 of 77
Quote:
Originally Posted by steelbom View Post
Is there a significant difference in audible noise from higher clocked memory? I.E, if you want a silent rig do you go for 1333MHz or is the difference between that and 1600 or 2133, or 2500 even, just not much?
Well, memory itself doesn't make noise (under normal circumstances). It's more about your cooling setup. Most performance memory (that we make, anyway) should operate at its rated spec with nothing more than the heat spreaders. That said, active cooling on your memory won't hurt anything.

If you really want quiet, get a good water-cooling setup.
Charity
(13 items)
 
  
CPUMotherboardGraphicsRAM
1090T Gigabyte GA880GMA-USB3 HD6870 1GB Changes, constantly. 
Hard DriveOptical DriveOSMonitor
1x300GB WD Velociraptor, 1x 1TB WD Caviar Generic Win7 Asus 24inch Wide Screen 
KeyboardPowerCaseMouse
Logitech Mushkin Volta 600W (MKNPS600-A) Lian-Li PC354B Logitech 
Mouse Pad
Seriously? 
  hide details  
Reply
Charity
(13 items)
 
  
CPUMotherboardGraphicsRAM
1090T Gigabyte GA880GMA-USB3 HD6870 1GB Changes, constantly. 
Hard DriveOptical DriveOSMonitor
1x300GB WD Velociraptor, 1x 1TB WD Caviar Generic Win7 Asus 24inch Wide Screen 
KeyboardPowerCaseMouse
Logitech Mushkin Volta 600W (MKNPS600-A) Lian-Li PC354B Logitech 
Mouse Pad
Seriously? 
  hide details  
Reply
post #53 of 77
Quote:
Originally Posted by DuckieHo View Post
I just want to point out your statement in factually incorrect.

A 3GHz CPU works at 0.333ns clock cycles.

The reason that RAM overclocking does not benefit is pretty complicated.... CPUs have more cache now for one... memory controller design.... other system bottlenecks.... application memory usage....

Memory revisions are also about power consumption, density, and signalling... not just clock speed.

It goes even further back... memory performance has not really been a bottleneck since post-Netburst.
Do you think Bulldozers 1866 MHz RAM will make much of a difference?

They also reportedly made improvements in speeds with the IMC...

I'm curious if the new Architect AMD has is going to produce much significance in performance with their faster RAM & IMC improvements...

Anyone?
The Rock
(15 items)
 
  
Reply
The Rock
(15 items)
 
  
Reply
post #54 of 77
The money spent on superfast ram is much better spent on a faster harddrive in almost every case. Hard drives have not advanced as quickly as other hardware components.

But fast ram is neat, and does have it uses.
Edited by pr0bie - 2/10/11 at 8:10pm
post #55 of 77
Quote:
Originally Posted by pr0bie View Post
The money spent on superfast ram is much better spent on a faster harddrive in almost every case. Hard drives have not advanced as quickly as other hardware components.

But fast ram is neat, and does have it uses.
SSDs are far from being awesome, they have quite a few issues too..

If they can make these significantly better..

Nevertheless, you are absolutely correct, faster data transfers from the drive is money better spent...

I paid $130 for DDR 1600 RAM back when the i5-750 first came out, and I was like man, that's too expensive for something that isn't much better than DDR2...

However, a SSD did increase performance significantly, despite the fact that they are very expensive..
Edited by _GTech - 2/10/11 at 8:14pm
The Rock
(15 items)
 
  
Reply
The Rock
(15 items)
 
  
Reply
post #56 of 77
Quote:
Originally Posted by _GTech View Post
Do you think Bulldozers 1866 MHz RAM will make much of a difference?

They also reportedly made improvements in speeds with the IMC...

I'm curious if the new Architect AMD has is going to produce much significance in performance with their faster RAM & IMC improvements...

Anyone?
I hope they've made significant improvements to the memory controller. AMD's memory bandwidth and controllers have really been lacking recently.
Bench Stuffs
(32 items)
 
ITX Daily Driver
(11 items)
 
For Sale: 6GHz+ 980X
$225 (USD) or best offer
CPUCPUCPUCPU
Intel Core i7 920 Intel Xeon W3520 Intel Core i7 4770K Intel Core 2 Duo E8500 
CPUCPUCPUCPU
Intel Pentium G3258 Intel QX9650 AMD FX 4300 Intel E2140 
CPUMotherboardGraphicsGraphics
E4300 Asus Rampage Extreme / P5E3 Deluxe / Gigabyte X... MSI GTX 275 Lightning MSI HD 7970 Lightning 
GraphicsGraphicsGraphicsGraphics
Asus GTX 285 Matrix Platinum MSI HD 5870 Lightning EVGA GTX 980 TI Classified K|ngp|n Edition MSI GTX 580 Lightning 
RAMRAMRAMRAM
Corsair Dominator Platinum 2666 10-12-12 Samsung Corsair Dominator GT 2000 7-8-7 Hypers Kingston HyperX 2000 9-9-9 BBSE G.Skill Trident 2000 9-9-9 PSC 
Hard DriveHard DriveHard DriveCooling
Kingston SSD Now! 128GB Patriot Torqx 2 32GB Seagate 80GB XPSC Rasa + EK VGA + HWLabs Black Ice GTX 360 +... 
OSOSOSKeyboard
W7 Pro x64 SP1 Windows XP x86 SP3 W8.1 Ducky Shine 3, Cherry Brown Switches 
PowerCaseMouseMouse Pad
Corsair AX1200 Dimastech Easy V2.5! <3 Logitech G5 Battlefield 2142 Edition Puretrak Stealth 
  hide details  
Reply
Bench Stuffs
(32 items)
 
ITX Daily Driver
(11 items)
 
For Sale: 6GHz+ 980X
$225 (USD) or best offer
CPUCPUCPUCPU
Intel Core i7 920 Intel Xeon W3520 Intel Core i7 4770K Intel Core 2 Duo E8500 
CPUCPUCPUCPU
Intel Pentium G3258 Intel QX9650 AMD FX 4300 Intel E2140 
CPUMotherboardGraphicsGraphics
E4300 Asus Rampage Extreme / P5E3 Deluxe / Gigabyte X... MSI GTX 275 Lightning MSI HD 7970 Lightning 
GraphicsGraphicsGraphicsGraphics
Asus GTX 285 Matrix Platinum MSI HD 5870 Lightning EVGA GTX 980 TI Classified K|ngp|n Edition MSI GTX 580 Lightning 
RAMRAMRAMRAM
Corsair Dominator Platinum 2666 10-12-12 Samsung Corsair Dominator GT 2000 7-8-7 Hypers Kingston HyperX 2000 9-9-9 BBSE G.Skill Trident 2000 9-9-9 PSC 
Hard DriveHard DriveHard DriveCooling
Kingston SSD Now! 128GB Patriot Torqx 2 32GB Seagate 80GB XPSC Rasa + EK VGA + HWLabs Black Ice GTX 360 +... 
OSOSOSKeyboard
W7 Pro x64 SP1 Windows XP x86 SP3 W8.1 Ducky Shine 3, Cherry Brown Switches 
PowerCaseMouseMouse Pad
Corsair AX1200 Dimastech Easy V2.5! <3 Logitech G5 Battlefield 2142 Edition Puretrak Stealth 
  hide details  
Reply
post #57 of 77
Quote:
Originally Posted by SgtSpike View Post
I'm dying to know... what do you do and what is the significant difference?

I bought 12GB DDR3 for $135.


I dunno man, on newegg right now, there's a dozen 2x2GB 1333mhz kits for $39.99, but the cheapest 2x2GB 1600mhz kit is $41.99. Where are you getting your ultra-cheap 1600mhz ram?
It's close enough, I think some other sites have cheaper as well. Plenty of the 1333Mhz generic kits sell for above $40 as well, above that it's a complete ripoff; unfortunately the majority of generic kits purchased are brand-based i.e. G.Skill Ripjaws, have heatspreaders, some even have unacceptable 1.65V ratings! RAM should never ever be judged based on brand unless taking support/warranty into account.

There is a Crucial 1333 CL7 kit for $40 right now (before the promo code), that's a pretty good one since it's got Micron ICs and may be very overclockable. Matches/beats the price of all generic kits.
Edited by xd_1771 - 2/10/11 at 8:25pm
post #58 of 77
Quote:
Originally Posted by wh-ATI View Post
...so the only reason to get high frequency RAM is e-peen.
High bandwidth low latency RAM is faster than the cheaper counter-part. It's most unfortunate no one understands RAM, and no one benchmarks it properly. This OP test is no exception, yet another website/review that does not understand how to benchmark memory properly, so they arrive at an incorrect conclusion, and spread even more misinformation to the community.

It's even more sad that they use a weak chipset with dual channel.. and not even fast memory modules. 2133 CL9 has a latency of 8.44ns, which is sad considering 2 year old Elpida modules have latency at sub-7ns, I've seen as low as 6.6ns. Maybe if this reviewer got his or her hands on some real memory, such as 1800MHz CL6, 2100MHz CL7, or 2400MHz CL8 and configured it properly, we could see some real results.
Edited by CorkyFan - 2/10/11 at 9:09pm
post #59 of 77
Quote:
Originally Posted by CorkyFan View Post
High bandwidth low latency RAM is faster than the cheaper counter-part. It's most unfortunate no one understands RAM, and no one benchmarks it properly. This OP test is no exception, yet another website/review that does not understand how to benchmark memory properly, so they arrive at an incorrect conclusion, and spread even more misinformation to the community.

It's even more sad that they use a weak chipset with dual channel.. and not even fast memory modules. 2133 CL9 has a latency of 8.44ns, which is sad considering 2 year old Elpida modules have latency at sub-7ns, I've seen as low as 6.6ns. Maybe if this reviewer got his or her hands on some real memory, such as 1800MHz CL6, 2100MHz CL7, or 2400MHz CL8 and configured it properly, we could see some real results.
But what does it gain you if you go from 8ns to 6ns? What real-world benefit do you see?
post #60 of 77
Quote:
Originally Posted by SgtSpike View Post
But what does it gain you if you go from 8ns to 6ns? What real-world benefit do you see?
None afaik.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [TR] Exploring the impact of memory speed on Sandy Bridge performance