Sempron 2800+ AMD\’s Phoenix

28 Jul, 2004 at 11:17pm ET Article published in Tech

Comments

1. 
   28 Jul 2004 ~ 11:28pm MediaMan Little background for you here. Sempron wasn't due to launch until August 18th. There was PLENTY of time to review this product. So I thought. 9AM July 26th I got notification that the launch had been bumped up a tad...to the 28th....OF JULY!!! 0 to 60 on a processor review in 48 hours?!?!
   
   
2. 
   28 Jul 2004 ~ 11:33pm primesuspect Nice work, doug!
3. 
   29 Jul 2004 ~ 12:20am mmonnin Thanks AMD for the review.
   
   Good job Doug!!
   
   Nice price on some of the low CPUs.
4. 
   29 Jul 2004 ~ 1:21am MediaMan Note a price discrepency on versus our information from AMD on the press release and the AMD website It's a 6 dollar difference across the board. I'll inquire.
5. 
   29 Jul 2004 ~ 10:21am GHoosdum Doug, I'm really impressed by your quick turnaround time on this.
   
   And I'm impressed by the Sempron's performance there...
6. 
   29 Jul 2004 ~ 11:25am MediaMan I was surprised too. There are, of course, different reviews out there with different results. Remember that across the reviews hardware configuration does change but one thing is clear...Sempron delivers a better bang for the buck compared to Celeron-D.
   
   I've yet to get full into the 3100+ and further bench the 2800+ for an even better analysis. There just wasn't time. The benchmark numbers still had the wet paint sign on as the review went on-line.
   
   We all knew what a good performer the Thorughbred B core was/is. Sempron is that core so while it may no longer wage a good battle at the top end against the top-of-the-line P4s...AMD has just re-assigned it to another front...the battle against Celeron-D.
7. 
   29 Jul 2004 ~ 4:55pm PressX Nice review... looks better than I thought.
8. 
   29 Jul 2004 ~ 8:30pm MediaMan Prices adjusted to reflect AMD Processor Pricing. For those of you who saw the other prices; distributor pricing. Not much of a markup...about $6.
9. 
   2 Aug 2004 ~ 8:03am Unregistered One of the worst reviews I've seen for a while. What was the point of using different graphics card (and even integrated ones!) in different setups and then trying to compare the results? This review tells absolutely nothing about the performance level of Sempron... Truly amateurish work.
10. 
    2 Aug 2004 ~ 8:42am GHoosdum

    Unregistered wrote:

    One of the worst reviews I've seen for a while. What was the point of using different graphics card (and even integrated ones!) in different setups and then trying to compare the results? This review tells absolutely nothing about the performance level of Sempron... Truly amateurish work.

    
    Your comment makes no sense. CPUs can be, and often are, benchmarked independently of the GPU all the time. Furthermore, the release of the Sempron was done early, and MM had to struggle to get the article done in two days' time - he admits it himself, and even states that the article will be rewritten once he gets some systems together for straight-up comparisons.
    
    There's no need to be so spiteful just because the P4 looks inferior... Do you work for Intel or something?
11. 
    2 Aug 2004 ~ 11:29am Dexter Great review under pressure, MediaMan. Ignore the unregistered comment as a drive-by insult: the person does not even have the decency to register as a user and tell us who he/she is. Apparently they don't understand the process of a review, which normally requires comparisonsa and benchmarks under different sets of circumstances.
    
    Dexter...
12. 
    2 Aug 2004 ~ 11:43am TheBaron

    Unregistered wrote:

    One of the worst reviews I've seen for a while. What was the point of using different graphics card (and even integrated ones!) in different setups and then trying to compare the results? This review tells absolutely nothing about the performance level of Sempron... Truly amateurish work.

    
    its a budget cpu, it damn well should be tested against budget graphics chipsets as well - I.E. integrated gfx
13. 
    2 Aug 2004 ~ 12:14pm Thrax

    Unregistered wrote:

    One of the worst reviews I've seen for a while. What was the point of using different graphics card (and even integrated ones!) in different setups and then trying to compare the results? This review tells absolutely nothing about the performance level of Sempron... Truly amateurish work.

    
    Truly amateurish method of stating your opinion. Anonymous trolls.
14. 
    2 Aug 2004 ~ 12:16pm dragonV8 I just laugh at people who hide behind a cloak of a anonymity. People like that make me look better.
    
    Good write up MM.
15. 
    2 Aug 2004 ~ 1:20pm Unregistered The unregistered commentator is back.
    
    Do you always register when you visit a site for a first time? I don't. It would just take too much time and effort to do in every site I visit (especially when there is no easy way to do it instantly). So it's not hiding, just pure common sence.
    
    And no, I don't work for intel. I actually am bit of a AMD fan.
    
    Anyway, why does this review suck? Well, it doesn't answer the question that this kinda review should always answer: How fast is the Sempron? There are just too many variables. Out of the 9 test, 8 measure the overall performance of the system, and only one pure proseccor performance.
    
    Ofcourse it's nice to find out the real life overall performance of system, but when compared systems differ from each other so dramatically (=different graphics cards), there just is no way to tell how much a proseccor effects the result, and what are the effects of the other parts used. What makes this review especially bad is that 8 out of 9 test had this mistake.
    
    This test might aswell been nForce2 IGP vs. Radeon 9100 IGP test (or actually it would have been more usefull that way). If you don't have time to do a proper review (which would tell something about Semprons speed) it's better to leave the review undone alltogether. Releasing this kinda review just makes me (and undoubtably some others aswell) avoid the site in the future...
16. 
    2 Aug 2004 ~ 8:44pm MediaMan Let me address a few points that the guest brought up.
    
    1) We allowed for unregistered comments to encourage people to comment. Some users hesitate at signing up for anything. I wanted to give site visitors more opportunity to comment be it good or bad.
    
    2) The "dragstrip of benchmarks"
    
    What is "real world performance?" It is a subjective interpretation. I always encourage any visitor to read our reviews and at least 2-3 other reviews from our colleague's sites. The big picture will give a far more accurate interpretation.
    
    Sempron is for a budget buyer. Those buyers typically do not purchase top-end video cards. They buy a "Dell" on a price point. Those PCs have low end graphic cards or integrated graphics cards. Budget buyers don't go out and spend $500 on a video card. They spend $500 on the whole system.
    
    Sempron is for that marketplace.
    
    I agree with the unregistered guest that the comparison wasn't perfect. The review had a full disclaimer at the first benchmark page to explain the tight turnaround and the reasons.
    
    There'll be a pause to explain the benchmarking procedure before the flames start firing in. Sempron wasn't due to launch until mid-August. Short-Media only found this out 48 hours prior to the new launch date of July 28th. 2004. Our goal was to get a budget buyer's perspective of Sempron to publication. It is important to put Sempron into the proper perspective when examining the slew of benchmarks from this and other publications.
    
    To that end there was very little time to benchmark Sempron let alone comparative systems. Two weeks would have been tight and two days was just silly. To that end we pulled data from our archives that was as close a comparison as possible.
    
    If the unregistered guest does not want to visit this site in the future due to this then I am sad to lose a visitor. Comments...even negative comments that are posted without malice or poor language are constructive feedback.
    
    The unregistered guest did bring up a very important point.
    
    How fast is the Sempron? There are just too many variables.
    
    The overall message isn't about how fast Sempron is. It's, as stated quite clearly, the value/performance equation. If you want "fast" then purchase the top end AMD or INTEL products. This is the Achilles Heal of reviews. It all comes down to the pretty graphs where a visitor can quickly, without wasting a neuron, see how "fast" one product is compared to another. Add up the pretty charts and then make a decision to buy or not.
    
    That isn't the best way to buy but it is what readers expect.
    
    The second part of the unregistered guest's comment was There are just too many variables.
    
    This is too true. All of us don't buy the same components. We mix and match. We overclock and we don't. We use different OS and software configurations.
    
    That is why I always endorse and push people to read our reviews and two to three others to get a better overall picture. I get the same hardware in as other review sites. Show me two or three out of the dozen review sites who got the same numbers...
    
    If the hardware is the same...shouldn't the numbers be?
    
    Again...I welcome the unregistered guest's comments; present and future...good an bad. If this particular guest thinks that this review "sucks" then so be it. If they do not visit again then I hate to lose a visitor and it reminds me to work all that much harder to keep future visitors coming back.
17. 
    2 Aug 2004 ~ 11:15pm Madball And how does it overclock? Are they locked?
18. 
    3 Aug 2004 ~ 1:15am Dexter

    Unregistered wrote:

    Out of the 9 test, 8 measure the overall performance of the system, and only one pure proseccor performance.

    
    Around here, we like to call them processors.
    
    If you are going to be so hypercritical of a review of a processor, it least learn how to spell it, dude
    
    Dexter...
19. 
    3 Aug 2004 ~ 1:23am Thrax You have to allow everyone a typo, Dexter. It's obvious from the rest of his post that he's rather literate.
20. 
    3 Aug 2004 ~ 2:18am Dexter

    Thrax wrote:

    You have to allow everyone a typo, Dexter. It's obvious from the rest of his post that he's rather literate.

    
    
    I "allowed" a typo: "commen sence."
    
    But "proseccor" typed twice in his comment...? Nah. Especially when the word was typed correctly 38 times in the article. If the dude read the article so closely that he was able to offer such incisive commentary on it, one would think he would have found the correct spelling of the item he seems to think he knows a lot about.
    
    Dexter...
21. 
    4 Aug 2004 ~ 3:47am Unregistered First of all it's kinda low to pick on my typos, especially considering that english is only my third language... Secondly, if you want to have a debate about something one of the basic rules is that stick to the subject (and don't get personal).
    
    But now I have to confess that before I posted my first comment, I didn't read the disclaimer. Instead I just browsed the article trough and got amazed about the conducted tests. And I would suspect that most people read reviews like I did; just browsing them quickly trough and checking the bechmarks (especially if you have red few Sepron reviwes earlier). So my initial reaction might not have been as justified as I tought it would be. Sorry for that...
    
    But still I still stand behind my two claims
    1. It's pointless to run gaming benchmarks in a processor review , when the limitating factor is the used graphics card. You might aswell used XP3200+ instead of Sempron, and the results would have been virtually same. Ofcourse you can still run couple of gaming test just to see how a typical budget system performs, but 8 out of 9 is still too much.
    
    2. If the article isn't good enough to be published (and you know it), don't publish it. Like I said, most just browse trough the article disregarding any disclaimers. If you want people to know that you have had the opportunity to review the product, then make a article focusing on something else than somewhat useless benchmarks.
    
    Hmmm... Perhaps I could register now
22. 
    4 Aug 2004 ~ 8:20am mmonnin MM just doesnt publish graphs for people like you to look at, he publishes something to read for the rest of us.
23. 
    4 Aug 2004 ~ 9:44am Straight_Man

    Unregistered wrote:

    One of the worst reviews I've seen for a while. What was the point of using different graphics card (and even integrated ones!) in different setups and then trying to compare the results? This review tells absolutely nothing about the performance level of Sempron... Truly amateurish work.

    
    Something to think about-- value boxes are likely NOT to have top line graphics. As a CPU in isolation article, or a CPU versus CPU-- pure CPU-- article, limiting graphics is maybe not the best approach. BUT, the Sempron is not likely to come on a hyper-high end system otherwise. So, it was real life as Sempron is expected to be bundled and is Targetted by AMD to be bundled-- it is intended to replace Duron and is priced accordingly. How many duron boxes come with the latest separate video card in system???
    
    For a 48 hour turn, VERY GOOD. As MM said, he will be digging into Sempron more-- expect some good digging with a practical and not theoretical perspective. This was NOT how to stretch to ub3rmost (that is largely ART and takes time to do right), this was a very good first look in a very short notice deadline timeframe.
    
    Unregistered one, please write a better review if you want.... Let us look at it, please. MM has lots of things he wants to write about, and does NOT do just this site. He's a career media person also. And practical. So he starts with practical scenarios. This is an enthusiast's site, not a pure engineering tech site. So, some folks come as techs, others (majority) come wanting to see what they can expect from a Sempron-centered SYSTEM.
    
    Over the coming months I would expect to see both from here (first looks plus detailed followups as MM stated would happen)-- not demanding, EXPECTING because of what has come before. EXPECTING GOOD, in other words. But CAREFUL, covering fine detail well, means you take time. Two days to test and write, you end up with a first look. Two-three weeks, you get and can take time to dig and try tricks and apply system art in configuring which the average bear in the woods does not know. Never should have been published??? US mags and sites for enthusiats do first look style things all the time.
24. 
    11 Aug 2004 ~ 10:18am Rogan

    1. It's pointless to run gaming benchmarks in a processor review , when the limitating factor is the used graphics card. You might aswell used XP3200+ instead of Sempron, and the results would have been virtually same. Ofcourse you can still run couple of gaming test just to see how a typical budget system performs, but 8 out of 9 is still too much.

    
    Utter Rubbish, all i wanted to know is how well sempron would run against an equivelent XP chip AND Intel. There is no reason why these should not be shown in benchmarks.
    
    Unfortunately the only part of your argument that was valid, that all graphics tests should be run using the same graphics card was lost because you trolled. Well played ... dumbass.
    
    Media Man, you gotta make sure that the GPU matches, I mean you can't always match up motherboards and memory, well you could between XP and Sempron in this case. But swapping graphics cards in a graphics benchmark nullifies your testing. I really wanted to see how the chips stacked up against each other, but as unregistered says (poorly) we cant see because you have swapped the hardware.
    
    Anyway, thx for the review.
25. 
    10 Sep 2004 ~ 1:19pm Unregistered < Not the same Unregistered guest >
    First, I'd like to complement MediaMan on taking the time to write the article in the first place. As a professional analyst, I know the time it takes the gather the material and to produce a readable piece for public consumption. Two days is indeed impressive.
    
    My main comment though, is that he should have considered using a Celeron based system (which are widely available) instead of a P4. As you have implied in your message, the Sepron is meant as a value alternative and the attraction to the platform will be its price/performance over equivalent Intel-based value systems. While it is impressive that the Sepron fares well against a higher priced P4 system, it would be more valuable to see, dollar for dollar, the incremental improvement of this system.
    
    And I agree with the other Unregistered guest, hammering someone over spelling errors are rather low.
    
26. 
    10 Sep 2004 ~ 2:16pm Thrax Actually, I'm of the opinion that if the Sempron holds its own against the Pentium 4, then it would do even better against the Celeron. That's the real test of value; budget versus enthusiast.
27. 
    10 Sep 2004 ~ 2:50pm MediaMan

    Unregistered wrote:

    < Not the same Unregistered guest >
    My main comment though, is that he should have considered using a Celeron based system (which are widely available) instead of a P4. As you have implied in your message, the Sepron is meant as a value alternative and the attraction to the platform will be its price/performance over equivalent Intel-based value systems. While it is impressive that the Sepron fares well against a higher priced P4 system, it would be more valuable to see, dollar for dollar, the incremental improvement of this system.

    
    
    You are most correct. The Sempron, since it is a (sort of) reclassified XP, should have been pitted against the Celeron. The reason I didn't was two-fold. 48 hour deadline and money.
    
    I was faced with purchasing a Celeron and motherboard to support it. (Some $500 CAD) Contrary to popular belief we don't make wads of cash so the bill would have been footed by me personally.
    
    If I had more than 48 hours then I would have done this. I could have sold off the Celery and partially recouped my investment. Making a site grow and giving the readers better articles does take that sort of committment. I have a 2" stack of receipts that show quite well that it isn't always "free stuff"
    
    I thank you for your commments. Everything helps me produce better articles for you. Glad that you did take the time to comment...It is appreciated.
    
    
    
    /me wanders off looking for his lottery tickets.
28. 
    10 Sep 2004 ~ 3:12pm Geeky1

    I have a 2" stack of receipts that show quite well that it isn't always "free stuff"

    
    Tell me about it... that XP-120 review ended up costing me like $350 (1 DFI LanParty, only to discover it didn't fit the XP-120, 1 ASUS P4P800E-Deluxe, 1 XP-120)
    
    But then, I got $350 worth of new toys to play with out of the deal too...
29. 
    10 Sep 2004 ~ 4:37pm Gobbles

    MediaMan wrote:

    You are most correct. The Sempron, since it is a (sort of) reclassified XP, should have been pitted against the Celeron. The reason I didn't was two-fold. 48 hour deadline and money.
    
    I was faced with purchasing a Celeron and motherboard to support it. (Some $500 CAD) Contrary to popular belief we don't make wads of cash so the bill would have been footed by me personally.
    
    If I had more than 48 hours then I would have done this. I could have sold off the Celery and partially recouped my investment. Making a site grow and giving the readers better articles does take that sort of committment. I have a 2" stack of receipts that show quite well that it isn't always "free stuff"
    
    I thank you for your commments. Everything helps me produce better articles for you. Glad that you did take the time to comment...It is appreciated.
    
    
    
    /me wanders off looking for his lottery tickets.

    
    
    I disagree. Using a P4 was just in my opinion as the celeron has never been able to compete with the duron/sempron...
30. 
    10 Sep 2004 ~ 7:00pm MediaMan

    Gobbles wrote:

    I disagree. Using a P4 was just in my opinion as the celeron has never been able to compete with the duron/sempron...

    
    
    Which is now how AMD has positioned Sempron so in the argument of which would you choose; Sempron or Celeron...which one would win and be the obvious choice?
    
    
31. 
    21 Jul 2005 ~ 12:03am no time to register why bicker instead of research? who cares which CPU plurbs the X axis fastest with the same equipment--within .0006MHz per nansecond per a pretty graph? this sort of review is pratical, a bit biased by the writer and the components, and very effective. excluding neophytes, one can read the review and form an opinion and have a starting point to dig. at this stage can't we all go to our chosen sites and get thespecs, again excluding neophytic box builders, the writer can offer an educated and hopefully reliable opinion on the value vs. horsepower of the CPU because the CPUs have different targets, determining to the Nth degree which is faster kind of skirts the usefulness of the article, doesn't it?
32. 
    21 Jul 2005 ~ 12:05am still too busy sorry for the typos befur innywun desides to plug me
33. 
    21 Jul 2005 ~ 12:15am primesuspect you're questioning the usefulness of an article that is almost a year old
    
    This processor, and any arguments surrounding it, are long obsolete.
34. 
    21 Jul 2005 ~ 12:04pm Nomad I want to see the written writ for exumming this thread's corpse :|
35. 
    21 Jul 2005 ~ 1:58pm profdlp

    Nomad wrote:

    I want to see the written writ for exumming this thread's corpse :|

    Ghoul.
36. 
    21 Jul 2005 ~ 4:13pm Gobbles

    MediaMan wrote:

    Which is now how AMD has positioned Sempron so in the argument of which would you choose; Sempron or Celeron...which one would win and be the obvious choice?
    
    

    
    sempron of course. I was saying that pitting a sempron against a p4 is a fair fight as a regular xp would have crushed a p4 and along that lines, the sempron would have crushed the celeron so bad all that would be left under the hsf would have been a small amount of silica dust.
    
    Im a tad back logged on my replies to threads...