perf #s for GRF vs 7500 Re: Anyone Deployed Ascend's GRF IP Switch?

["joseph j. kim" <jokim () ipcom com>]

On Tue, 26 Aug 1997, Joe  Shaw wrote:

> On Mon, 25 Aug 1997, joseph j. kim wrote:

> > From Cisco's:
>
>  technology preview of Cisco's new family of gigabit switch routers (GSRs)
>  providing high performance solutions ranging from 5 to 60 Gb/s for
>  Internet and large-scale WAN Intranet backbone applications.
>
> Now, I've got a GRF 400 with the ATM OC-3c card, the FDDI card, and the
> 10/100Base-T card.  It works flawlessly.  And I can't honestly see how the 
>
> > Also, not that I know anything about the GRF but I think Cisco
> > claims that 7500 real-world performance is much better than the GRF400.
>
> I've had cisco sales reps claim even more ludicrous things while the
> support engineer started turning red, so who knows...

hi,

Ok. since no one else bothered to post real numbers here are some:

Cisco numbers:
=============
                                Cisco 7500              Ascend GRF

Performance                     2 Gbps                  4 Gbps
System Bandwidth                1.4 Mpps                2.8 Mpps
Theoretical 

Performance with                65 - 85 Kpps            35 - 70 Kpps
180 byte                        (Bi-directional)        (Bi-directional)
packets

Line Card Forwarding            325 Kpps                280 Kpps
Rate                            (7507 w/ 5 VIP.s)       (Fully Loaded)
Real Performance                

Performance w/ Services         880 - 1000 Kpps         140 - 210 Kpps

Routing Table Size              250,000 +               150,000 +


The tolly/ascend report numbers:
===============================

looking at their data (n.t. = not tested):

# of modules    cisco 7514 w/rsp4       GRF1600         GRF400
---------------------------------------------------------------
1               122,300                 60,388          59,731
2               244,520                 120,848         119,348
3               366,516                 n.t.            n.t.
4               368,575                 241,516         236,776

6               367,774                 360,000
8               367,302                 483,016
16              n.t.                    965,424



Using Random IP- in this test the routers were re-booted and the
performance measured at 15 minutes after boot time. also, destination IP
addresses were varied by randomly generating class c dest. addresses.

random ip test (at +15 min. after boot):
# of modules    cisco 7513 w/rsp4       GRF1600
------------------------------------------------
1               41,096                  54,454
2               84,504                  107,720
3               72,762                  162,750
4               86,654                  217,700
5               71,867                  265,500


So, who's numbers should we believe or feel are more appropriate to real
world situations?

> Comparing GateD to IOS becomes more of a religious preference than
> anything else.  I'm content knowing both, truth be told.
>
> > maybe someone can post some performance numbers.

-jjk