AMD's Winter Update: Athlon II X3 455, Phenom II X2 565 and Phenom II X6 1100T
by Anand Lal Shimpi on December 7, 2010 12:01 AM EST- Posted in
- CPUs
- AMD
- Phenom II X6
- Athlon II
- Phenom II
x264 HD Video Encoding Performance
Graysky's x264 HD test uses x264 to encode a 4Mbps 720p MPEG-2 source. The focus here is on quality rather than speed, thus the benchmark uses a 2-pass encode and reports the average frame rate in each pass.
Video encoding and other thread heavy tasks are best suited for AMD's more-is-better core strategy. You get six cores on the 1100T and three with the 455, in both cases the competing Intel part doesn't stand a chance. In the second encoding pass the Athlon II X3 is over 30% faster than the Pentium G6950. Without unlocking additional cores, the Phenom II X2 565 BE doesn't impress here.
PAR2 Multithreaded Archive Recovery Performance
Par2 is an application used for reconstructing downloaded archives. It can generate parity data from a given archive and later use it to recover the archive
Chuchusoft took the source code of par2cmdline 0.4 and parallelized it using Intel’s Threading Building Blocks 2.1. The result is a version of par2cmdline that can spawn multiple threads to repair par2 archives. For this test we took a 708MB archive, corrupted nearly 60MB of it, and used the multithreaded par2cmdline to recover it. The scores reported are the repair and recover time in seconds.
The Phenom II X6 is competitive in our Par2 test, the Athlon II X3 455 is significantly faster than the Pentium G6950 and the Phenom II X2 565 falls short of its target. Rinse and repeat.
7-Zip Benchmark Performance
Included in 7-zip is a pure algorithm test that completely removes IO from the equation. This test scales with core count and as a result we get a good theoretical picture of how these chips perform. Note that the actual 7-zip compression/decompression process is limited to 2 threads so there's no real world advantage to having more cores.
65 Comments
View All Comments
vol7ron - Tuesday, December 7, 2010 - link
"today we're getting speed bumps"Usually this term is associated with a slow down, not a push forward - you slow down at the speed bump. It's sort of like how you don't associate a stop sign with accelerating as fast as you can, which many do right after.
Other Thoughts: Perhaps you meant speed burst?
fic2 - Tuesday, December 7, 2010 - link
Apparently you don't know that words can have multiple meanings.See #5 from http://www.thefreedictionary.com/bump
bump (bmp)
v. bumped, bump·ing, bumps
v.tr.
1. To strike or collide with.
2. To cause to knock against an obstacle.
3.
a. To knock to a new position; shift: bumped the crate out of the way.
b. To shake up and down; jolt: bumped the child on her knee; was bumped about on a rough flight.
4.
a. To displace from a position within a group or organization.
b. To deprive (a passenger) of a reserved seat because of overbooking.
5. To raise; boost: bump up the price of gasoline.
6. Sports To pass (a volleyball) by redirecting it with the forearms.
jabber - Tuesday, December 7, 2010 - link
We shall see. If Intel has been 100% successful and reliable in one aspect of computing its dissapointing with its GPU performance.Always a lot of bluster and pre-release pomp about how it will be many times better than the previous piece of crap and then it hits the floor like a dead moose dropped from 50 feet.
I dont see that changing radically.
kwantor - Tuesday, December 7, 2010 - link
How does that work exactly?On 2560x yes, you should get 100+fps on 1024x.
MrSpadge - Tuesday, December 7, 2010 - link
They're using "Ultra CPU" settings to stress the CPUs. People are probably not really running the game like this.MrS
nitrousoxide - Tuesday, December 7, 2010 - link
It will be bulldozed by eight-core SNB-E, but surely it will bulldoze quad-core SNBs.flyck - Wednesday, December 8, 2010 - link
lets wait and see shall we. If each BD core is faster then each K8 core it might be very close to it.MrSpadge - Tuesday, December 7, 2010 - link
Not sure about the US, but here in Germany the i7 860 has effectively been replaced by the i7 870 since a couple of months. It's 1 - 2 multipliers faster for a small price premium. Currently the 1100T is actually priced a hair above the 870.Given the approxiamte tie in threaded apps and wins for the 870 in lightly threaded apps and power consumption at any load level I'd certainly go with that one.
MrS
dertechie - Tuesday, December 7, 2010 - link
They're comparing to i7 860 since they already have one on hand to play with, I suspect.mapesdhs - Tuesday, December 7, 2010 - link
MrSpadge is right; when I went looking for one recently, I noticed many
stores had the 860 priced higher than the 870, or the gap was so small
that the 860 made no sense. So I bought an 870 instead.
Btw, it's a bit misleading IMO to include AMD oc results and yet not at
least briefly mention how well the Intel chips also oc, especially given earlier
reviews here of the i3 and other options, eg. I get 6.88 for Cinebench 11 with
my oc'd 870, 20442 for Cinebench 10 (and this isn't on the high side either).
I've found the 870 to oc better than my older 860 aswell, and not just because
of the base clock difference. It just seems to work better. I'm sure I could push
it to 4.5+, but there's no need for this on a gaming rig with two GTX 460s SLI.
Indeed, so far I find game fps scores are better with HT turned off and a
higher CPU clock, so even more headroom is possible (confirmed this effect
with 3DMark06, Unigine Heaven, Stalker COP and X3TC so far). With the
same Vcore/VTT, my 870 was ok at 4444 instead of 4270 with HT off, and
temps were lower. If you want max 3DMark06 overall scores, leave HT on;
to max out game fps rates though, try turning HT off and increase the raw
clock (I'd like to know which games benefits from HT - none of my current
tests show any gain).
Hmm, the '3dsmax 9 - SPECapc' test looks interesting, might give that a spin.
Is that a separate download to Viewperf 11?
Ian.