Does Bandwidth/Timings mean Anything?

[eBoy0]

I posted this on EOCF aswell, wanna hear your feedback guys, cancled my order of ballistix after seeing this.
http://forums.anandtech.com/messagev...&enterthread=y
Test Configuration:

-Video card is BFG6800GT@ultra.
-CPU Winch 939 1800@2610 Mhz, 290Mhz HTT, 3x LDT
-Operating system Windows 2000 service pack 4 and all updates.
-Dx version- 9c
-Lastest *official* Nvidia video card and Nvidia motherboard drivers.
-All memory is 2 x 512 MB.

Test 1: Super budget crucial PC2100 ram Henceforth refered to as “old ram”
290 bus
100 memsetting yeilds PC2300 (145Mhz)
3 3 3 1T

Test2: PC3200 Buffalo Value ram w/ CH-5 chips Henceforth refered to as “value ram”
290 Bus
133 mem setting Yeilds ~PC3000 (186Mhz)
3-3-3 1T

Test 3: Low Latency (LL) and PC3000, many ram can do this!! I used the TCCD here, but "value" buffalo can handle this speed and latency too!!
290bus
133 mem setting Yeilds ~PC3000 (186Mhz)
2-2-2 1T

Test 4: Highend: PQI PC3200 w/ Samsung TCCD
290Bus 1:1
200 memsetting Yeilds ~PC4700 (290Mhz)
3-4-3 1T

Test 5: Highend: Crucial Ballistix PC3200 (uses Micron G)
290Bus
166 memsetting Yeilds ~PC3800 (237Mhz)
2.5-2-2 1T

Testing software:

All real world except Sandra mem/cpu just to show how erronious it is for A64:

-SiSoftware Sandra Lite 2005, mem and CPU
-Far Cry 1.3 HardwareOC Far Cry Benchmark Demo , max details, AF = 1, AA = 0, 800x600 and 1600 x 1200
-Unreal Tournament 2003 @ 10x7 default settings
-Doom3- All High Quality Settings, "timedemo demo1 1", AA4x, 1600x1200 and 1024x768
-Half Life 2- All High Quality Settings, Route Kanal custom demo features indoor/outdoor, AA6x, AF4x, 1600x1200 and 1024x768
-LAME 3.96.1 (build from mitiok.cjb.net)Default settings 163MB CBS Masterworks Beethoven Symph #9 op.125 Allegro non troppo, un poco maestoso 16:44m
-Cinebench 2003
-ScienceMark 2.0 beta (23SEP03 build) - Molecular Dynamics
-PiFast Parameters: 0,0,10m,1024,1
-Superpi 1M
-Winrar 3.41 internal benchmark
-POV-Ray v3.6 Chess2.pov 1024x768 AA.03
-DVDShrink 3.2 Lawrence Of Arabia Superbit Encode Chap 1-6, 863MB to 500 MB = 58% compression (sharp quality setting)

Results

Sandr mem INT ALU
PC2300 - 3-3-3 1T- 4285
PC3000 - 3-3-3 1T- 5556
PC3000 - 2-2-2 1T- 5602

PC4700 - 3-4-3 1T- 7272
PC3800 - 2.5-2-2-1T - 6977

Sandra mem FLOAT FPU
PC2300 - 3-3-3 1T- 4288
PC3000 - 3-3-3 1T- 5550
PC3000 - 2-2-2 1T- 5603
PC4700 - 3-4-3 1T- 7267
PC3800 - 2.5-2-2-1T - 6904
As you can see, TCCD Sanda mem is showing a whopping ~70% faster vs. the cheapest PC2100 ram ever sold. And 30% more than value ram, PC3000 ram @3-3-3. Will this translate into real performance gains? Well lets see.

Sandra CPU Drystone ALU
PC2300 - 3-3-3 1T- 11064
PC3000 - 3-3-3 1T- 11068
PC3000 - 2-2-2 1T- 11066
PC4700 - 3-4-3 1T- 11071
PC3800 - 2.5-2-2-1T - 11068
no difference

Sandra CPU FPU
PC2300 - 3-3-3 1T- 4253
PC3000 - 3-3-3 1T- 4253
PC3000 - 2-2-2 1T- 4253
PC4700 - 3-4-3 1T- 4252
PC3800 - 2.5-2-2-1T -4353
no difference

Sandra CPU SSE2
PC2300 - 3-3-3 1T- 5527
PC3000 - 3-3-3 1T- 5536
PC3000 - 2-2-2 1T- 5528
PC4700 - 3-4-3 1T- 5551
PC3800 - 2.5-2-2-1T - 5446
meh, now lets get to the real world:

Farcry 800x600 in FPS
PC2300 - 3-3-3 1T- 61.46
PC3000 - 3-3-3 1T- 66.49
PC3000 - 2-2-2 1T- 72.67
PC4700 - 3-4-3 1T- 74.97
PC3800 - 2.5-2-2-1T - 77.08

Farcry 1600x1200 in FPS
PC2300 - 3-3-3 1T- 62.25
PC3000 - 3-3-3 1T- 68.14
PC3000 - 2-2-2 1T- 71.31
PC4700 - 3-4-3 1T- 73.52
PC3800 - 2.5-2-2-1T - 74.42
On farcry, we see ~18% faster performance from the old ram. But only ~8% faster than value ram. hint: spend the xtra $100 on a better video card. Notice async low latency Ballistix owning all even 1:1 TCCD.


Doom3 1024x768 in FPS
PC2300 - 3-3-3 1T- 77.8
PC3000 - 3-3-3 1T- 80.3
PC3000 - 2-2-2 1T- 81.4
PC4700 - 3-4-3 1T- 82.3
PC3800 - 2.5-2-2-1T - 82.4
No signifigant difference above PC3000

Doom3 1600x1200 in FPS
PC2300 - 3-3-3 1T- 44.6
PC3000 - 3-3-3 1T- 44.6
PC3000 - 2-2-2 1T- 44.6
PC4700 - 3-4-3 1T- 44.5
PC3800 - 2.5-2-2-1T - 44.6
Video card limitation here...

Half-life 2 1024x768 in FPS
PC2300 - 3-3-3 1T- 113.3
PC3000 - 3-3-3 1T- 120.9
PC3000 - 2-2-2 1T- 131.6
PC4700 - 3-4-3 1T- 133.3
PC3800 - 2.5-2-2-1T - 134.4
We see low timings here credited with speed since PC3000LL is basically equivalent to PC4700. Overall nothing big above PC 3000.

Half-life 2 1600x1200 in FPS
PC2300 - 3-3-3 1T- 98.77
PC3000 - 3-3-3 1T- 102.5
PC3000 - 2-2-2 1T- 106.0
PC4700 - 3-4-3 1T- 107.9
PC3800 - 2.5-2-2-1T - 107.3
Same as above. Save your money.

Pifast in seconds
PC2300 - 3-3-3 1T- 38.91
PC3000 - 3-3-3 1T- 36.53
PC3000 - 2-2-2 1T- 35.48
PC4700 - 3-4-3 1T- 34.67
PC3800 - 2.5-2-2-1T - 34.45
We see bandwitdh have *some* effect here..nothing to write home about. PC2300 vs PC4700 is 10.9% faster, but very small difference vs. value ram.

Superpi 1M in seconds
PC2300 - 3-3-3 1T- 37
PC3000 - 3-3-3 1T- 34
PC3000 - 2-2-2 1T- 33
PC4700 - 3-4-3 1T- 32
PC3800 - 2.5-2-2-1T -32
Similar as above. Nothing drastic at anything above PC3000

Science Mark- Molecular Dynamics time in seconds
PC2300 - 3-3-3 1T- 67.03
PC3000 - 3-3-3 1T- 66.92
PC3000 - 2-2-2 1T- 66.68
PC4700 - 3-4-3 1T- 66.69
PC3800 - 2.5-2-2-1T - 66.47
Seems bandwidth has no effect here.

CineBench 2003 Rendering
PC2300 - 3-3-3 1T- 363
PC3000 - 3-3-3 1T- 365
PC3000 - 2-2-2 1T- 366
PC4700 - 3-4-3 1T- 368
PC3800 - 2.5-2-2-1T - 367
Seems bandwidth has no effect here.

CineBench 2003 Shading
PC2300 - 3-3-3 1T- 414
PC3000 - 3-3-3 1T- 426
PC3000 - 2-2-2 1T- 434
PC4700 - 3-4-3 1T- 439
PC3800 - 2.5-2-2-1T - 440
Infinitesmal bandwidth help

CineBench 2003 OpenGL SW
PC2300 - 3-3-3 1T- 1949
PC3000 - 3-3-3 1T- 1992
PC3000 - 2-2-2 1T- 2019
PC4700 - 3-4-3 1T- 2032
PC3800 - 2.5-2-2-1T - 2034
Infinitesmal bandwidth help

CineBench 2003 OpenGL HW
PC2300 - 3-3-3 1T- 3568
PC3000 - 3-3-3 1T- 3662
PC3000 - 2-2-2 1T- 3756
PC4700 - 3-4-3 1T- 3781
PC3800 - 2.5-2-2-1T - 3796
this is getting old, on to another game.

UT2003 10x7 Flyby
PC2300 - 3-3-3 1T- 297
PC3000 - 3-3-3 1T- 319
PC3000 - 2-2-2 1T- 334
PC4700 - 3-4-3 1T- 341
PC3800 - 2.5-2-2-1T - 345
14% max boost, hey but very small difference with value ram.

UT2003 10x7 BotMatch
PC2300 - 3-3-3 1T- 110
PC3000 - 3-3-3 1T- 118
PC3000 - 2-2-2 1T- 123
PC4700 - 3-4-3 1T- 127
PC3800 - 2.5-2-2-1T - 128
15% boost over PC2300, only 8% over value ram.

Winrar in KB/s
PC2300 - 3-3-3 1T- 496
PC3000 - 3-3-3 1T- 564
PC3000 - 2-2-2 1T- 636
PC4700 - 3-4-3 1T- 651
PC3800 - 2.5-2-2-1T - 703!!!
Here we see pretty signifigant gains due to bandwidth 41% diff top to bottom. This test obviously is very beneficial to have bandwidth or LL. Update: Crucial Ballistix owning all, Winrar seems the like LL even more than bandwidth. We sorta saw this already by the very close scores of PC3000 LL VS PC4700 high bandwitdh.

Lame in seconds
PC2300 - 3-3-3 1T- 56
PC3000 - 3-3-3 1T- 56
PC3000 - 2-2-2 1T- 56
PC4700 - 3-4-3 1T- 56
PC3800 - 2.5-2-2-1T -56
Seems bandwidth or LL has no effect here.

Chess2.pov in seconds
PC2300 - 3-3-3 1T- 246
PC3000 - 3-3-3 1T- 245
PC3000 - 2-2-2 1T- 246
PC4700 - 3-4-3 1T- 247
PC3800 - 2.5-2-2-1T -246
Seems bandwidth or LL has no effect here.

DVDShrink in seconds
PC2300 - 3-3-3 1T- 243
PC3000 - 3-3-3 1T- 235
PC3000 - 2-2-2 1T- 228
PC4700 - 3-4-3 1T- 221
PC3800 - 2.5-2-2-1T -224
DVDShrink seems to like bandwidth a little bit.


Summary of Results
Well there it was. While sandra likes to show whopping 70% increased bandwidth of expensive TCCD it really does'nt translate to real world . Only one area seemed signifigant was winrar. The couple game tests which slow slight inprovement with bandwidth or LL, I think you're much better of upgrading to the next tier of video card if on a budget instead of investing with overpriced ram. Investing in expensive Ram probably offers to poorest bang for the buck of any computer part. What's intesting was Ballistix's async 166 and LL the performance faired a little better than TCCD 1:1. Seems A64 prefers LL rather than bandwidth? Also shows that running async is'nt to be shunned.

Course if you want the best, go right ahead and invest in expensive ram, as shown it makes a slight difference in gaming and winrar but I'm sticking with $130 2x512 value ram and use a divider from now on because the price to performance ratio is simply not worth it. IMO You're much better off using that cash for the next level of video card which really does boost performance by ~70% in games.

One other thing of intrest is PC3000 running async 133 at low timings 2-2-2 was almost as good as PC4700 high bandwidth but loose timings TCCD is forced to run at 290Mhz.... many a ram can do ~200Mhz 2-2-2.


Bottom Line:

Some Games = very slight benefit to using expensive high bandwidth TCCD or expensive LL ram Like Ballistix over value ram. Winrar sees major benefit though, 41% top to bottom.

LAME, Cinebench, ScienceMark, PiFast, Superpi, POV-Ray, DVDShrink shows no or infinitesmal boost above Value ram.

[targ]

nice write up. thanks.

you obviously put some thought into this and it shows, like you said, thats ram gives little real world performance increase.

targ

[eBoy0]

nice write up. thanks.

you obviously put some thought into this and it shows, like you said, thats ram gives little real world performance increase.

targ

I'd love to take the credit, but it's not mine i just posted the words from the review, lemme make it more obvious.

[eBoy0]

I would say that good ram lets you OC better tho all around but if you are running stock than it seems that cheap ram will for sure get you by.

But you see, if running a divider doesn't hinder performance... and since there is no performance increase in bandwidth.. meaning there is no show of performance increase between 200mhz and 300mhz, and by this article, timings don't seem to much increase in appz, why not just get like 2x1gig of the loosest/cheapest memory?

[Ugly n Grey]

running a divider on AMD makes no difference as there is no such thing as 1:1 on an integrated memory controller, it's ALWAYS on a divider, the BIOS setting just let's you change what that divider is.

Extremely helpful post eBoy0, I bet more peeps will spend that extra on the vid card instead of the RAM after reading this.

On a separate note, I think hard multitasking on multiple cores/procs requires the bandwidth as you are getting double the hit if you are working both procs on memory intensive applications. Multi threaded SQL server or Oracle for example can use all the memory bandwidth and hard drive speed you can give it and still ask for more.....

On a single proc? More often there's some other performance bottle neck besides memory. Get the good video

Great post.
UnG

[ben805]

Yup, for my X2 I'll take as much memory bandwidth as I possibly could. I probably would careless if I'm running single core.

[snowwie]

no in the scheme of things they mean very little. but in xtremesystems they mean a bit more.

[eBoy0]

running a divider on AMD makes no difference as there is no such thing as 1:1 on an integrated memory controller, it's ALWAYS on a divider, the BIOS setting just let's you change what that divider is.

Extremely helpful post eBoy0, I bet more peeps will spend that extra on the vid card instead of the RAM after reading this.

On a separate note, I think hard multitasking on multiple cores/procs requires the bandwidth as you are getting double the hit if you are working both procs on memory intensive applications. Multi threaded SQL server or Oracle for example can use all the memory bandwidth and hard drive speed you can give it and still ask for more.....

On a single proc? More often there's some other performance bottle neck besides memory. Get the good video

Great post.
UnG

Thanks, having you said this.. i'm wondering what 2gig kit i should get now, i cancled my ballistix just a while back after reading this. Any preferences UnG?

[Ugly n Grey]

My personal preferences are HIGHLY biased, and not always on the technology side. I'll explain myself a lil

For registered ECC RAM I've always bought from Crucial, they test well and that means I never have hassles. It's just plug and play in my Opty systems and I've had two RMA's in four years that were handled very efficiently. That's somewhere on the order of 300 GB of RAM. I buy a lot of Tyan boards and they all seem to work well.

For my desktop style systems I've bought pretty much everything in the last three years from OCZ. Same reason as above, geat customer service, no hassle assistance and returns, manufacturer support is great and they have a decent selection of products from "el cheapo" to "el emptyo my wallet". I will likely keep buying from them and my latest 2 GB kits have been OCZ OCZ4002048PFDC-K for stock systems and I picked up set of OCZ5002048EBPE-K off a friend to play with. I'm still using 4x512's in my own box....I didn't notice anything from 2T performance wise while I was actually using the computer.

But I've seen some great deals on 2Gb kits and I wouldn't hesitate to shop around and get something that fits my budget normally, it's just that I'm a real stickler for support and when I find what I like I never change. That's why I always buy a T series Thinkpad for example, they basically fall all over themselves to help you out, after all you paid for it in the premium price.

In other words, those are my choices but I'm not the best person to ask....

[eBoy0]

My personal preferences are HIGHLY biased, and not always on the technology side. I'll explain myself a lil

For registered ECC RAM I've always bought from Crucial, they test well and that means I never have hassles. It's just plug and play in my Opty systems and I've had two RMA's in four years that were handled very efficiently. That's somewhere on the order of 300 GB of RAM. I buy a lot of Tyan boards and they all seem to work well.

For my desktop style systems I've bought pretty much everything in the last three years from OCZ. Same reason as above, geat customer service, no hassle assistance and returns, manufacturer support is great and they have a decent selection of products from "el cheapo" to "el emptyo my wallet". I will likely keep buying from them and my latest 2 GB kits have been OCZ OCZ4002048PFDC-K for stock systems and I picked up set of OCZ5002048EBPE-K off a friend to play with. I'm still using 4x512's in my own box....I didn't notice anything from 2T performance wise while I was actually using the computer.

But I've seen some great deals on 2Gb kits and I wouldn't hesitate to shop around and get something that fits my budget normally, it's just that I'm a real stickler for support and when I find what I like I never change. That's why I always buy a T series Thinkpad for example, they basically fall all over themselves to help you out, after all you paid for it in the premium price.

In other words, those are my choices but I'm not the best person to ask....

Alright dude, thanks for your honesty. I guess i'll have to search around or.... be lazy and make a thread on it seperately.

[Vapor]

I would have liked to have seen the effect of 261 2-2-2-0 and 290 2.5-3-3-0 in that bunch as well....oh well.

While the gains aren't as great from going to RAM twice the price (usually), it's still a gain.

[iboomalot]

Granted the Ballistix might only be 8% faster ,, a Raptor in Raid is only 10% faster than a 7200 rpm HHD ,, the 7800GTX is only 25% faster than a X800XT but all those small gains in every aspect do add up to a faster machine but that speed does get costly.

for a faily inexpensive setup might try this

168.00 amd 2800+ skt 754 + Gigabyte GA-K8VM800M combo
180.00 (4) 512 sticks of Geil Value
65.00 Maxtor 160GB 7200RPM 8MB SATA/150 Hard Drive
140.00 Nvidia 6600GT
100.00 19" CRT

at 653.00 not the cheapest system but large 19" screen + 2 gig of ram and decent video card won't be cheap then again that entire setup will be around the cost of a new X1800XT video card

[MikalCarbine]

7800GTX is only 25% faster than a X800XT

Is it only that much??!

[jahjahbinks]

What about Pi score? Does higher bandwidth give a better result than tighter timing?

[Vapor]

Read the first post

[jahjahbinks]

lol, i completely missed it. my bad.

[iboomalot]

Is it only that much??!

only in 3d2005 in farcry and UT2004 the ATI is only behind a few frames or less than 7%

either way 7-25% increase between a 300ish card vs a 500ish priced card.

very hard to get large gains in the computer business without alot of extra cost.

yes ballistix and OCZ and many more 2gig sticks are 300+ in cost but just a yr ago what would you have paid for 2gig???

nice info but I wouldn't include sandra CPU tests thats a waste of time and space showing CPU tests when you are trying to compare RAM.

otherwise nice info and seems on par.

[Repoman]

Yeah, I saw that a while ago too. I would have gone for something cheap but I saw something else about dual core and memory bandwidth, I know single core gains nothing over like 5GB/s, so theoretically with DC you need all you can get until 10GB/s.

Anyway, they ran moderate multitasking (don't exactly remember what) but it was 2 timed memory intensive tasks using slow memory and fast memory on separate cores. They gained like 20 sec on the faster memory, it was something like 80 vs 100 seconds. Those aren't exact and I have no idea where I saw that, but it's true.

[Ugly n Grey]

When it come to memory intensive tasks, a pair of single core Opterons with Numa are WAY faster on memory intensive tasks on a 64 bit OS. I gain 70% performance increase over a dual core on some DB tasks... There's something to be said for two memory controllers....but considering most games aren't built for the 64 bit OS I can't see it making any diff at all in that environment.

[targ]

what i would like to see was the same experiment on a Fx-57 @ 3.6 or so.

as its obvious, that 2.6Ghz only NEEDS PC3000, as after that there is only very small increases. this could well be that the processor cant deal with any more data, so what if the clock speed is increased? how with the bandwidth dependancy change.

that or some one could do it at 2.0Ghz or something to see if the bandwidth requirement goes down.

targ

[gbomb944]

It's pretty much like everything else. The best thing to get is something that has value as in is fairly cheap that you can oc a little and then get close to premium performance for not so premium a price. Do a little googling and find out what 2gig kits can oc some and still are reasonable. Also check 2x1024 oc database.

I've been doing some research on the same question and have figured out what I'm going to get. It's about 60$ cheaper than balistics and has better timing plus 20-30% oc ability. So 20% cheaper and oc's by 30% at better timings = value and midrange price.

[Zebo]

Why Cancel when Ballistix is only $52 at mwave for 512MB? I wrote this when highend ram was $250+ a gig and you could get cheap ram for less than half.

[LIONS*LEADER]

i may have misunderstood.

its my understanding that (on intel) the processor fsb is directly resposible for the ability of higher frequency memory to perform.


so say you have a 800fsb processor that isnt overclocked. the performance gain of ddr533 3:4 , at the same timings as ddr400 1:1 , is very minimal

therefor any higher frequency than the processor fsb can handle would show very little gain. it may even hurt it if the memory must loosen the timings in order to achive the higher frequencies.



however.


if you had 1066fsb processor the performance gain of ddr533 1:1 , at the same timings as ddr400 4:3, is very signifigant.




its my understanding that this is why the HTTx4 multi is the highest performing when overclocking since you can run the memory at the same bandwidth as the processors hyper transport (FSB)

.

[Exlua]

You get bad PC2100 Crucials my Crucials do >235FSB @ 2,5-2-2-2-11 @3V