Memory Latency Impact on Performance
We just looked at the impact of memory bandwidth on performance, but what about latency? Let's first by adjusting the CAS latency from our default of 2 clocks up to 3 clocks. Almost all DDR400 these days is CAS 2 memory, but older memory may have a higher CAS latency or you may have to increase your CAS latency when overclocking to gain more memory bandwidth, so what kind of a performance hit is there when going from CAS 2 to CAS 3?
at_canals_08 |
at_coast_05 |
at_coast_12 |
at_prison_05 |
at_c17_12 |
|
Tcl = 2 |
116.12 |
140.43 |
123.37 |
113.69 |
83.15 |
Tcl = 3 |
115.52 |
137.07 |
121.91 |
113.37 |
79.92 |
At worst, CAS 2 memory seems to be about 5% faster than CAS 3 memory when looking at at_c17_12, our most CPU intensive test. While 5% alone isn't anything major, combine that with a number of other performance tweaks and they can definitely begin to add up.
Now let's look at keeping Tcl (CAS latency) fixed at 2 clocks, but vary Trcd timings from 3 up to 6 clocks:
at_canals_08 |
at_coast_05 |
at_coast_12 |
at_prison_05 |
at_c17_12 |
|
Trcd = 2 |
116.12 |
140.43 |
123.37 |
113.69 |
83.15 |
Trcd = 3 |
115.71 |
136.99 |
122.46 |
113.08 |
79.97 |
Trcd = 4 |
113.92 |
134.42 |
120.87 |
112.38 |
79.83 |
Trcd = 5 |
113.42 |
131.82 |
119.34 |
114.79 |
79.12 |
Trcd = 6 |
113.23 |
128.26 |
117.56 |
111.15 |
77.4 |
For the most part we saw no real changes when adjusting Trcd, the one exception being at_coast_05 which actually showed a pretty big difference between a Trcd value of 2 and higher latency values.
Next we'll look at adjusting Trp:
at_canals_08 |
at_coast_05 |
at_coast_12 |
at_prison_05 |
at_c17_12 |
|
Trp = 2 |
116.12 |
140.43 |
123.37 |
113.69 |
83.15 |
Trp = 3 |
115.6 |
139.24 |
123.13 |
116.35 |
82.09 |
Trp = 4 |
115.85 |
138.88 |
122.98 |
113.16 |
82.05 |
Trp = 5 |
114.84 |
138 |
122.65 |
112 |
80.98 |
Trp = 6 |
114.5 |
136.95 |
121.96 |
115.61 |
80.95 |
Here we see very little impact on performance.
Putting them all together we can see what the overall impact on using fast DDR400, higher latency DDR400 and extremely high latency DDR400 will be:
at_canals_08 |
at_coast_05 |
at_coast_12 |
at_prison_05 |
at_c17_12 |
|
2-2-2-10 |
116.12 |
140.43 |
123.37 |
113.69 |
83.15 |
3-3-3-10 |
114.47 |
134.11 |
120.64 |
112.62 |
80.56 |
3-6-6-10 |
110.74 |
123.76 |
114.75 |
112.17 |
73.8 |
Our standard 2-2-2-10 memory does actually offer reasonable performance benefits in Half Life 2 compared to DDR400 with higher timings such as 3-3-3-10 or the unrealistically high 3-6-6-10.
First and foremost Half Life 2 does appear to be rather dependent on memory bandwidth, but it is also quite appreciative of low latency memory as well. If you're wondering whether being able to run memory at low timings and high clock speeds is important, when it comes to Half Life 2 performance it is.
Facebook
Twitter
RSS
68 Comments
View All Comments
Phantronius - Wednesday, January 26, 2005 - link
#3NF4 will not be supporting AGP bud, sorry, its PCI-E from here on out.
Ozz1113 - Wednesday, January 26, 2005 - link
Ill backup the thought of putting some T-bred cores in there. My OC'd XP2600 333 w\ modded 9500 radeon system ran HL2 very well. I would have liked it to have been better, but it is not worth upgrading yet.Araemo - Wednesday, January 26, 2005 - link
Hehehe.. I'll fifteenth the "Please include an AXP3200+" sentiment. Personally, I'd rather see one or two AXPs included than a complete list of athlon64s.. You can generally extrapolate the performance of a given CPU if you are given two other CPUs with the same cache/FSB/core. I know that my Mobile barton handled the game fine, but I'd like to know how far behind a cheap A64 it really is.miketheidiot - Wednesday, January 26, 2005 - link
"Our standard 2-2-2-10 memory does actually offer reasonable performance benefits in Half Life 2 compared to DDR400 with higher timings such as 3-3-3-10 or the unrealistically high 3-6-6-10."Reasonable performance benefits? decent 2.5-3-3-10 ram can be found cheap nowdays (http://www.newegg.com/app/ViewProductDesc.asp?desc... and compared to the oczrev2 and other $200+ modules is at least $60 cheaper, in some case as much as $100 cheaper. The 2-2-2 is only 2% faster than the 3-3-3, so does that extra $60+ really offer "reasonable performance benefits"?
Jeff7181 - Wednesday, January 26, 2005 - link
#13... I wouldn't call that an error, I'd call that a difference in opinion. :)PrinceGaz - Wednesday, January 26, 2005 - link
There seems to be an error on page 4- "Almost all DDR400 these days is CAS 2 memory, but older memory may have a higher CAS latency..."Shouldn't it say "Almost all DDR400 tested by AnandTech is premium CAS 2 memory, but CAS 2.5 and CAS 3 are more common..."
Aquila76 - Wednesday, January 26, 2005 - link
What about SLI configs? I think people looking SLI for an option may want a better idea how their CPU choice affects the dual GPU choice. Can you add SLI'd 6600GT, 6800GT, 6800 Ultra benches to the tables at the end of your article?Tiamat - Wednesday, January 26, 2005 - link
Yeah, although probably unrealistic, tossing AXP's into a "low end range" comparison along side would help some people. Overall, great article!Regs - Wednesday, January 26, 2005 - link
I liked the Ram latency and 64-bit/128-bit test. But I'm wondering how would 2.5 Cas would perform? Makes sense to list it since a lot of value named brand ram modules come with 2.5 CAS. I would think it would perform in-between the two, but I'm having the slightest inkling that 2.5 CAS and 2.0 CAS will perform the same.Can't wait to hear about multithreaded games for dual core CPUs.
Jeff7181 - Wednesday, January 26, 2005 - link
Well we can look at the other HL2 articles cause there's an XP3200 in those... but... this being a CPU oriented article I thought it would be nice to have that CPU included. Possibly even an XP3000 so we can get an idea for how it scales so I can estimate how my 2.48 GHz Mobile AXP compares.