New XMOS chips??!

Technical discussions around xCORE processors (e.g. General Purpose (L/G), xCORE-USB, xCORE-Analog, xCORE-XA).
User avatar
mon2
XCore Legend
Posts: 1912
Joined: Thu Jun 10, 2010 11:43 am

Post by mon2 »

It appears that only a gifted few have received the boards - without charge. So far, I would agree with those that have received the new kit as they are quite valuable to this user forum - the same have assisted where the vendor has not. Support in general is quite weak and on a need to know basis and if lucky will follow weeks after opening a ticket. From our last dialog, the kits should be available through at least Digikey sometime this month or May. Very sad that a 3rd person has to inform you of these details when it is the sole responsibility of the chip vendor. On this note, do review the documentation in detail for the related new parts as there continues to be typos that need attention. For example, some BGA balls are marked NC in the datasheet, yet they are used in the schematic. This "simple" typo can lead to thousands of dollars of damage.

As a long time paying customer for all such development boards for years, the support needs to be kicked up a few notches. Personally, not a way I would run my company. Still strongly feel that XMOS has potential with our company as we have invested years into the R&D with these devices. However, the support continues to be consistently weak.

Mon2


User avatar
segher
XCore Expert
Posts: 844
Joined: Sun Jul 11, 2010 1:31 am

Post by segher »

The datasheets say PRELIMINARY all over, and that is a
good thing ;-)

I think the announcement was rushed a bit because a
distributor spilled the beans. Give it some time?
Hagrid
Active Member
Posts: 44
Joined: Mon Jul 29, 2013 4:33 am

Post by Hagrid »

I am at a key point in my product development where I have specific issues to solve. The new chips are a solution. The new Explorer board is very near perfect as a platform for my product as I am nudging the limits on the startkit's memory.

At a "build or buy" decision point I can only move forward with more info. Price is the main datum closely followed by availability.

The teasing should stop with some facts so I (and others) can make some commercial decisions.

As Johnny 5 said, "Need input!"
Redeye
XCore Addict
Posts: 129
Joined: Wed Aug 03, 2011 9:13 am

Post by Redeye »

mon2 wrote:It appears that only a gifted few have received the boards - without charge. So far, I would agree with those that have received the new kit as they are quite valuable to this user forum - the same have assisted where the vendor has not. Support in general is quite weak and on a need to know basis and if lucky will follow weeks after opening a ticket. From our last dialog, the kits should be available through at least Digikey sometime this month or May. Very sad that a 3rd person has to inform you of these details when it is the sole responsibility of the chip vendor. On this note, do review the documentation in detail for the related new parts as there continues to be typos that need attention. For example, some BGA balls are marked NC in the datasheet, yet they are used in the schematic. This "simple" typo can lead to thousands of dollars of damage.

As a long time paying customer for all such development boards for years, the support needs to be kicked up a few notches. Personally, not a way I would run my company. Still strongly feel that XMOS has potential with our company as we have invested years into the R&D with these devices. However, the support continues to be consistently weak.

Mon2

I totally agree with everything in this post. As another customer that has several years of R&D invested in XMOS the support does vary between weak and completely non-existent. The point about potential datasheet errors is particularly valid - I have previously been the victim of a datasheet error which cost me thousands of dollars and several weeks of wasted R&D effort to track down. The response from XMOS when I pointed it out? *shrug*. I find it both reassuring that I'm not alone in experiencing the dreadful support and worrying that it's a more widespread problem that XMOS aren't doing anything about.

So, these new chips leave me with a bit of a dilemma for my next product platform - go through the same development cycle of pain without any help from XMOS and re-use most of my existing code, or cut my losses and switch processors to a supplier who offers proper product documentation and support?
User avatar
Folknology
XCore Legend
Posts: 1274
Joined: Thu Dec 10, 2009 10:20 pm

Post by Folknology »

Redeye wrote:
mon2 wrote:It appears that only a gifted few have received the boards - without charge. So far, I would agree with those that have received the new kit as they are quite valuable to this user forum - the same have assisted where the vendor has not. Support in general is quite weak and on a need to know basis and if lucky will follow weeks after opening a ticket. From our last dialog, the kits should be available through at least Digikey sometime this month or May. Very sad that a 3rd person has to inform you of these details when it is the sole responsibility of the chip vendor. On this note, do review the documentation in detail for the related new parts as there continues to be typos that need attention. For example, some BGA balls are marked NC in the datasheet, yet they are used in the schematic. This "simple" typo can lead to thousands of dollars of damage.

As a long time paying customer for all such development boards for years, the support needs to be kicked up a few notches. Personally, not a way I would run my company. Still strongly feel that XMOS has potential with our company as we have invested years into the R&D with these devices. However, the support continues to be consistently weak.

Mon2

I totally agree with everything in this post. As another customer that has several years of R&D invested in XMOS the support does vary between weak and completely non-existent. The point about potential datasheet errors is particularly valid - I have previously been the victim of a datasheet error which cost me thousands of dollars and several weeks of wasted R&D effort to track down. The response from XMOS when I pointed it out? *shrug*. I find it both reassuring that I'm not alone in experiencing the dreadful support and worrying that it's a more widespread problem that XMOS aren't doing anything about.

So, these new chips leave me with a bit of a dilemma for my next product platform - go through the same development cycle of pain without any help from XMOS and re-use most of my existing code, or cut my losses and switch processors to a supplier who offers proper product documentation and support?
Whilst I agree that in the past some of the support has been left wanting around documentation/erata, I believe that Xmos have significantly improved this over recent years and months. For example you can now actually subscribe to any of their documentation (datasheets, application notes etc..) which enables you to stay completely informed on any documentation changes.

Also as has been mentioned before the new XXX-200 series datasheets are clearly marked as preliminary on every page and should be treated as such given that these are not even shipping products (as is the eXplorer 200 kit).

regards
Al
Redeye
XCore Addict
Posts: 129
Joined: Wed Aug 03, 2011 9:13 am

Post by Redeye »

Folknology wrote:Whilst I agree that in the past some of the support has been left wanting around documentation/erata, I believe that Xmos have significantly improved this over recent years and months. For example you can now actually subscribe to any of their documentation (datasheets, application notes etc..) which enables you to stay completely informed on any documentation changes.

Also as has been mentioned before the new XXX-200 series datasheets are clearly marked as preliminary on every page and should be treated as such given that these are not even shipping products (as is the eXplorer 200 kit).

regards
Al
True, the irony being that it was by following a change in pinout labeling on an updated datasheet that I got from a subscribed update that left me with a load of boards with L1 chips that wouldn't boot. That they can change a datasheet from a working pinout to one that can never possibly work is pretty frightening and really erodes trust in both the company and product.

I take your point that these datasheets are marked as preliminary, but I have experienced a great deal of wasted time and money from XMOS getting fully released datasheets seriously wrong and more importantly doing virtually nothing to help/support me.
User avatar
Ross
XCore Expert
Posts: 768
Joined: Thu Dec 10, 2009 9:20 pm

Post by Ross »

Redeye wrote: more importantly doing virtually nothing to help/support me.
I wasn't involved in this, and I won't be getting involved in any more replies but I think this statement is more than a little unfair, certainly quite a bit of XMOS time was devoted to tracking this down including inspecting your boards.

Let's try and keep threads on topic please.
User avatar
segher
XCore Expert
Posts: 844
Joined: Sun Jul 11, 2010 1:31 am

Post by segher »

On topic! What a nice idea :-)

Some observations that might or might not be correct:

- The "dual issue" mode is much more like a VLIW mode
than like a superscalar RISC; instructions are paired in the
instruction stream, in your code if you want, and a 32-bit
"bundle" is issued every cycle. Add nop fillers where needed.
The second instruction in a bundle can have the register
that the first instruction writes as input, and then it gets
the old contents of that register. A bit like exposed pipeline
machines.

- It seems that instruction fetches from ROM can be executed
at the same time as data accesses to RAM? Very nice :-)
User avatar
lilltroll
XCore Expert
Posts: 956
Joined: Fri Dec 11, 2009 3:53 am
Location: Sweden, Eskilstuna

Post by lilltroll »

Hagrid wrote:I am at a key point in my product development where I have specific issues to solve. The new chips are a solution. The new Explorer board is very near perfect as a platform for my product as I am nudging the limits on the startkit's memory.

At a "build or buy" decision point I can only move forward with more info. Price is the main datum closely followed by availability.

The teasing should stop with some facts so I (and others) can make some commercial decisions.

As Johnny 5 said, "Need input!"
Have you tried to contact XMOS sales directly?
User avatar
leon_heller
XCore Expert
Posts: 546
Joined: Thu Dec 10, 2009 10:41 pm
Location: St. Leonards-on-Sea, E. Sussex, UK.

Post by leon_heller »

Were Redeye's problems discussed on this forum?