ARM Challenging Intel in the Server Market: An Overview
by Johan De Gelas on December 16, 2014 10:00 AM ESTOverview of the Competitors
Let's sum everything up in one big table.
ARM/Intel SoC 2015 Comparison | ||||||||
SoC | Intel Xeon-D | Intel Atom C2000 | AppliedMicro X-Gene 1 (X-Gene 2) |
AMD A1100 | Cavium Thunder-X | Broadcom Vulcan | ||
Architecture | Broadwell | Silvermont | Storm (ShadowCat) | A57 | Thunder-X | Vulcan | ||
Cores Socket |
8 single |
8 single |
8 (16) sngle |
4-8 single |
16-48 dual |
20? | ||
Max. CPU Clockspeed | GHz | 2.4GHz | 2.4GHz (2.8GHz) |
2GHz | 2.5 Ghz | 3GHz | ||
Process technology | Intel 14nm | Intel 22nm | TSMC 40nm (TSMC 28nm) |
GF 28nm | GF 28nm | TSMC 16nm | ||
L1 Cache | 32KB I 32KB D |
32KB I 24KB D |
32KB I (*) 32KB D (*) |
48KB I 32KB D |
78KB I 32KB D |
32KB I 32KB D |
||
Decode | 4 | 2 | 4 | 3 | 2 | 4 | ||
Max. IPC (int) | 5 | 2 | 4 | 3 | 2 | 4 | ||
Exe Ports | 8 | 4 | 8 | 8 | 4? | 6 | ||
Max. FP Performance | 2x 256 bit | 1x 128 bit | 2x 128 bit | 2x 128 bit | 2x 128 bit | 2x 128 bit | ||
OoO buffer | 192 | 32 | >100 | 128 | 40 | 180 | ||
L2 Cache | 8x 256KB | 4x 1MB | 4x 256KB? (*) | 4x 1MB | 16MB | 20x 256KB | ||
L3 Cache | 8MB? | - | 8MB | 8MB | - | ? | ||
Max. RAM | 128GB | 64GB | 128GB | 128GB | 1TB | ? | ||
Memory Bus Width | 2x 64-bit | 2 x 64-bit | 4x 64-bit | 2x 64-bit | 4x 64-bit | 4x 64-bit | ||
DRAM (best) | DDR4- 2133 |
DDR3- 1600 |
DDR3- 1866 |
DDR3- 1866 |
DDR4- 2133 |
DDR4- 2133 |
||
TDP (top SKU) | 45W | 20W | 40W (25 W?) |
25W | +/- 95 W | ? | ||
Available | Q2-Q3 2015 |
Early 2014 |
Now (Q2 2015?) |
Q1-Q2 2015 |
Q1 2015 |
Q3 2015 |
(*) Deduced from Ganesh's article about the Helix SoCs
These are paper specifications of course, so they should be interpreted with a grain of salt. It looks like the AMD A1100 should top the Atom C2000 and go after the low end of the Xeon E3. AMD's Opteron A1100 is already available, but the current development kits do not hit the clock speed and performance targets.
The Thunder-X single-threaded performance in "traditional workloads" might only be at the level of the Atom C2000, but scale-out and network/crypto acceleration could give some remarkable results in certain workloads. The Cavium SoC is the hardest to predict and will show a very variable performance profile as it also incorporates many very specialized hardware accelerators. The Thunder-X reference servers are announced and should be available in the coming weeks.
The X-Gene is currently the widest ARM architecture with extra hardware acceleration mostly focused on networking. The X-Gene TDP was great on paper (25W when announced) but there are many indications (40W TDP) that AppliedMicro really needs the 28nm X-Gene 2 to be truly competitive in the performance/watt battle arena. The X-Gene 2 should be available around Q2 2015.
78 Comments
View All Comments
aryonoco - Wednesday, December 17, 2014 - link
I just wanted to thank you Johan De Gelas for this very insightful and interesting article.Hugely enjoyed reading it and your thoughts on the subject.
Good to see high quality content continue to be published at AT now that Anand has left.
JohanAnandtech - Wednesday, December 17, 2014 - link
aryonoco, Jann Thanks for letting me know. A good motivation to always push a bit harder to make sure I don't let my readers down :-).jann5s - Wednesday, December 17, 2014 - link
Thank you Johan, for writing this very interesting article!przemo_li - Wednesday, December 17, 2014 - link
Very well written walk through current and possible CPU/SOC parts.Will there be similar piece for software?
ARM (embedded) folks aren't famous for quality drivers/code.
It must change, so it will change. But for now such overview would be great!
bobbozzo - Wednesday, December 17, 2014 - link
Typo on page2:"(4 Slots x 8 DIMMs)" - change 8 to 8GB
Thanks
bobbozzo - Wednesday, December 17, 2014 - link
and page 4:"you will be able to choose between SoCs that have 100 Gbit Ethernet and 10GBit Ethernet."
should 100 be 40?
bobbozzo - Wednesday, December 17, 2014 - link
Page 12:"Most of them are the usual IPSec, TPC offloading engines"
Should that be TCP?
Also, are there still accelerators for AntiVirus engines and IDS/IPS search (there were some back in 2005).
Thanks
bobbozzo - Wednesday, December 17, 2014 - link
...I guess that's what the RegEx would be useful for.
However, not all IDS/IPS / A/V patterns use RegEx, and there are other means of acceleration.
eanazag - Wednesday, December 17, 2014 - link
Welcome back Johan.Glad to see you're still writing here. Good stuff in the article.
JKflipflop98 - Wednesday, December 17, 2014 - link
I simply don't get where this whole "microserver" thing is coming from.By the time you cluster up enough ARM processors to match the processing power of an Intel/AMD solution, you're burning just as much power and spent just as much money as you would have by using x86 in the first place. Except now you have to use some janky middleware solution because all your software is x86 and you're running on ARM cores.